{ "schema_version": 4, "ranking_kind": "capability_scale", "ranking_policy": "direct_upstream", "details": { "claude-opus-5-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "releasedAt": "2026-09-22", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "inputCny": 26.8092, "outputCny": 134.046, "cachedCny": 1.3404600000000002, "officialUrl": "https://platform.claude.com/docs/en/models/opus-5-5/overview", "note": null }, "overall": { "rank": 1, "score": 73.9, "pendingReason": null, "onBoard": true, "coverage": 0.964, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 1, "score": 74.3, "coverage": 0.857, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 2, "score": 76.8, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 2, "score": 71.3, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 4, "score": 72.7, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 74.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "18.5%", "reading": 107.5, "bound": null, "weight": "capped", "fallback": 0.12, "sourceRank": 1, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "107.5" }, { "label": "回退", "display": "12%" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "65.2%", "reading": 82, "bound": null, "weight": "full", "fallback": 0.1111111111111111, "sourceRank": 1, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "82.0" }, { "label": "回退", "display": "11%" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "59.6%", "reading": 75.5, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "89.3%", "reading": 75.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "claude-opus-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "75.1" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "30.4%", "reading": 74, "bound": null, "weight": "full", "fallback": 0.16, "sourceRank": 1, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "74.0" }, { "label": "回退", "display": "16%" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "90.3%", "reading": 72.4, "bound": null, "weight": "full", "fallback": 0.08, "sourceRank": 5, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.4" }, { "label": "回退", "display": "8%" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "66.9%", "reading": 71.8, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.8" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "capped", "fallback": 0.017857142857142856, "sourceRank": 14, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" }, { "label": "回退", "display": "2%" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 76.8, "items": [ { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "59.0%", "reading": 85.5, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 3, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "85.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "47.1%", "reading": 82.8, "bound": null, "weight": "full", "fallback": 0.02857142857142857, "sourceRank": 3, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "82.8" }, { "label": "回退", "display": "3%" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "71.0%", "reading": 81.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "claude-opus-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-29T17:34:54.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "81.6" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "91.2%", "reading": 79.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "claude-opus-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-22T18:27:39.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "79.8" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "49.5%", "reading": 78.2, "bound": null, "weight": "full", "fallback": 0.0045045045045045045, "sourceRank": 2, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "78.2" }, { "label": "回退", "display": "不到 1%" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "61.4%", "reading": 75.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.7" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "31.7%", "reading": 73.1, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.1" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "92.2%", "reading": 70.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "claude-opus-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 71.3, "items": [ { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,807", "reading": 80.6, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "80.6" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "73.5%", "reading": 80.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "80.4" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,866.3", "reading": 76.6, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "76.6" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "55.0%", "reading": 76.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "76.2" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "56.3%", "reading": 74.4, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 4, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "74.4" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "69.5%", "reading": 71.8, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 3, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.8" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "80.3%", "reading": 68.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-opus-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "58.6%", "reading": 66.3, "bound": null, "weight": "full", "fallback": 0.012592592592592593, "sourceRank": 9, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.3" }, { "label": "回退", "display": "1%" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "49.8%", "reading": 61.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 18, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "3.8%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "anthropic/claude-opus-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 72.7, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "66.7%", "reading": 97.1, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 3, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "97.1" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "46.4", "reading": 83.3, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "83.3" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "72.2%", "reading": 83.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-opus-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-22T18:27:39.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "83.2" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "86.3%", "reading": 72.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "claude-opus-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "72.8" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "62.0%", "reading": 70.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "anthropic/claude-opus-5.5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "70.5" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "84.7%", "reading": 63.8, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 5, "sourceModelName": "claude-opus-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.8" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "65.7%", "reading": 55.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 27, "sourceModelName": "claude-opus-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 72.1, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,503.7", "reading": 69.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-opus-5.5-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "69.9" } ] } ] } ], "excluded": [ { "key": "vals-migration", "name": "Code Migration", "reason": "评测方公布这次运行有 82% 的题目被转给了同厂商的旧型号，超过一半，不代表这个型号。" } ], "unmeasured": [ { "key": "epoch-chess", "name": "Chess Puzzles" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-opus-5-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-opus-5-5", "fetched_at": "2026-10-09T04:16:59.438Z", "raw_sha256": "0c060e884e11ed3b7bee4a09b2b36340ecc4a7ba31ba79248f62eecd6ff5dfa6" } }, "gemini-4-argon": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "releasedAt": "2026-09-30", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": null, "weightsUrl": null }, "price": null, "overall": { "rank": 2, "score": 73.5, "pendingReason": null, "onBoard": true, "coverage": 0.679, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 3, "score": 71.7, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 4, "score": 74.7, "coverage": 0.571, "sourceCount": 4, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 1, "score": 76.4, "coverage": 0.75, "sourceCount": 7, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 5, "score": 72.4, "coverage": 0.4, "sourceCount": 2, "onBoard": true } ], "metricCount": 21, "evidence": [ { "key": "coding", "name": "编程", "score": 71.7, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 7, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "57.6%", "reading": 78.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "78.3" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "68.2%", "reading": 76.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "57.1%", "reading": 74.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "74.6" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "91.9%", "reading": 74.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "74.2" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "61.8%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 74.7, "items": [ { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "44.3%", "reading": 81.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "81.7" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "45.5%", "reading": 75.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "75.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "57.1%", "reading": 72, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.0" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "27.1%", "reading": 70.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.3" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 76.4, "items": [ { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "82.4%", "reading": 90.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "90.7" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "19.6%", "reading": 85.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "85.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "77.5%", "reading": 76.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "76.7" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "65.4%", "reading": 73.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "73.8" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "58.8%", "reading": 72.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "google/gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,489.9", "reading": 67.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.7" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,626.2", "reading": 67.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.1" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 72.4, "items": [ { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "42.4", "reading": 80, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "80.0" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.7%", "reading": 54.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 73, "sourceModelName": "gemini-4-argon", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.8" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 75.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,525.2", "reading": 77.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-4-argon-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "77.8" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "livebench-coding", "name": "LiveBench · 编程综合" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-4-argon", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-4-argon", "fetched_at": "2026-10-09T04:16:58.897Z", "raw_sha256": "fce5788afdf6dbf331ec46cb4a1572206208ea1a738e01f4b16082b7f9a3cb00" } }, "claude-fable-5-1": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "releasedAt": "2026-09-01", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "inputCny": 67.023, "outputCny": 335.115, "cachedCny": 1.675575, "officialUrl": "https://platform.claude.com/docs/en/models/fable-5-1/overview", "note": null }, "overall": { "rank": 3, "score": 72.3, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 4, "score": 71.7, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 6, "score": 74.2, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 4, "score": 70.2, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 2, "score": 74.5, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 36, "evidence": [ { "key": "coding", "name": "编程", "score": 71.7, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "7.0%", "reading": 89.8, "bound": null, "weight": "capped", "fallback": 0.165, "sourceRank": 2, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "89.8" }, { "label": "回退", "display": "17%" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "58.1%", "reading": 78.5, "bound": null, "weight": "full", "fallback": 0.1111111111111111, "sourceRank": 4, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "78.5" }, { "label": "回退", "display": "11%" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "52.0%", "reading": 72.9, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 14, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.9" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "90.3%", "reading": 72.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.4" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "28.0%", "reading": 72, "bound": null, "weight": "full", "fallback": 0.28, "sourceRank": 3, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.0" }, { "label": "回退", "display": "28%" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "86.4%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-fable-5-1-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "54.6%", "reading": 69, "bound": null, "weight": "full", "fallback": 0.015384615384615385, "sourceRank": 9, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.0" }, { "label": "回退", "display": "2%" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "63.1%", "reading": 66.5, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 3, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": 0.03571428571428571, "sourceRank": 5, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" }, { "label": "回退", "display": "4%" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 74.2, "items": [ { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "40.0%", "reading": 79.9, "bound": null, "weight": "full", "fallback": 0.05714285714285714, "sourceRank": 6, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "79.9" }, { "label": "回退", "display": "6%" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "43.3%", "reading": 78.9, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 8, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "78.9" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "90.2%", "reading": 78.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-fable-5-1_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-01T18:38:24.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "78.3" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "47.7%", "reading": 77.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "77.2" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "58.0%", "reading": 74.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-fable-5-1_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-01T18:45:06.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "74.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "59.1%", "reading": 73.8, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.8" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "29.7%", "reading": 71.9, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 16, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.9" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "91.7%", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-fable-5-1-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "47.0%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "claude-fable-5-1_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-01T18:32:00.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 70.2, "items": [ { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "68.6%", "reading": 75.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-fable-5.1", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "75.8" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "54.0%", "reading": 75.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "claude-fable-5.1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "75.3" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "57.5%", "reading": 75.3, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.3" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,674.6", "reading": 75.2, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 6, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.2" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,756.2", "reading": 72.2, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 4, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.2" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "80.3%", "reading": 68.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "claude-fable-5-1-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "59.4%", "reading": 66.6, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 36, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.6" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "58.9%", "reading": 66.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.6" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "53.5%", "reading": 65.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.9" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "6.7%", "reading": 62.1, "bound": null, "weight": "full", "fallback": 0.016666666666666666, "sourceRank": 26, "sourceModelName": "anthropic/claude-fable-5-1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" }, { "label": "回退", "display": "2%" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 74.5, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "71.1%", "reading": 101.5, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "101.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "70.8%", "reading": 81.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-fable-5-1_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-01T18:31:59.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "81.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "89.5%", "reading": 81.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-fable-5-1-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "81.3" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "43.5", "reading": 80.8, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 3, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "80.8" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "64.0%", "reading": 72.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "anthropic/claude-fable-5.1", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "72.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "73.0%", "reading": 66.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-fable-5-1-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "66.3" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "85.3%", "reading": 65.1, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 4, "sourceModelName": "claude-fable-5-1", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.1" }, { "label": "回退", "display": "含回退" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 71.1, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,501.1", "reading": 68.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-fable-5.1-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "68.9" } ] } ] } ], "excluded": [], "unmeasured": [], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-fable-5-1", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-fable-5-1", "fetched_at": "2026-10-09T04:16:59.552Z", "raw_sha256": "21ae705a7937af21ffa08bed4619293250737857e7b5b2612a72193ddcabe989" } }, "gpt-6-astra": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "releasedAt": "2026-09-03", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "inputCny": 67.023, "outputCny": 335.115, "cachedCny": 6.7023, "officialUrl": "https://developers.openai.com/api/docs/models/gpt-6-astra", "note": null }, "overall": { "rank": 4, "score": 71.3, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 5, "score": 69.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 1, "score": 77.6, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 8, "score": 67.2, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 1, "score": 74.8, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 36, "evidence": [ { "key": "coding", "name": "编程", "score": 69.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "5.5%", "reading": 85.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 4, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "85.6" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "59.6%", "reading": 79.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "79.3" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "67.7%", "reading": 76.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.2" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "59.1%", "reading": 75.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.3" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "89.6%", "reading": 71.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "71.7" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "27.6%", "reading": 71.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "71.7" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "80.4%", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "56.5%", "reading": 57.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 32, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "82.1%", "reading": 57.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 30, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 77.6, "items": [ { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "84.0%", "reading": 90.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-30T14:57:43.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "90.2" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "62.9%", "reading": 89.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "89.2" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "72.0%", "reading": 87.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-30T14:57:43.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "87.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "63.3%", "reading": 87.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "87.4" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "93.7%", "reading": 84, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-30T14:57:43.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "84.0" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "53.2%", "reading": 80.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "80.3" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "31.7%", "reading": 73.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "92.7%", "reading": 71.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "71.8" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "54.7%", "reading": 69.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 67.2, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "83.0%", "reading": 73, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "73.0" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "64.7%", "reading": 72.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "72.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "68.5%", "reading": 71.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.2" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,568.9", "reading": 70.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.9" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "51.2%", "reading": 70.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "46.7%", "reading": 69.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "69.2" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,541.9", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.5%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "48.5%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "5.4%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 74.8, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "75.6%", "reading": 87.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6-astra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-30T14:57:43.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "87.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "89.4%", "reading": 81.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "81.1" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "43.4", "reading": 80.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "80.8" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "69.0%", "reading": 78.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "openai/gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "78.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.6%", "reading": 70.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.5" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "35.0%", "reading": 69.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "80.7%", "reading": 56.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 54, "sourceModelName": "gpt-6-astra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 67.2, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,477.1", "reading": 60.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "gpt-6-astra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "60.1" } ] } ] } ], "excluded": [], "unmeasured": [], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-6-astra", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-6-astra", "fetched_at": "2026-10-09T04:16:58.879Z", "raw_sha256": "eab49674c9aa7a5f0eaf445ec782396de261ab3b53332ae0085f4ffe95ecd1da" } }, "claude-sonnet-5-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "releasedAt": "2026-09-28", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "inputCny": 13.4046, "outputCny": 67.023, "cachedCny": 1.3404600000000002, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 5, "score": 71.3, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 2, "score": 73.9, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 5, "score": 74.3, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 3, "score": 70.6, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 15, "score": 63.4, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 73.9, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "6.5%", "reading": 88.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 3, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "88.5" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "64.1%", "reading": 81.5, "bound": null, "weight": "full", "fallback": 0.020202020202020204, "sourceRank": 2, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "81.5" }, { "label": "回退", "display": "2%" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "91.4%", "reading": 80.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-sonnet-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "80.1" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "69.8%", "reading": 77.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "77.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "63.6%", "reading": 76.9, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "76.9" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "92.4%", "reading": 74.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "74.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": 0.03571428571428571, "sourceRank": 5, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" }, { "label": "回退", "display": "4%" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "61.0%", "reading": 63.7, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 5, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.7" }, { "label": "回退", "display": "含回退" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 74.3, "items": [ { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "53.3%", "reading": 83.1, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 5, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "83.1" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "45.7%", "reading": 82.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "82.3" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "65.0%", "reading": 78.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "claude-sonnet-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-29T15:15:14.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "78.3" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "49.1%", "reading": 78, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "78.0" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "88.8%", "reading": 76.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-sonnet-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-29T15:15:14.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "76.5" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "31.4%", "reading": 73, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 7, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.0" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "55.0%", "reading": 70.2, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 9, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.2" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "91.6%", "reading": 69.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-sonnet-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.5" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 70.6, "items": [ { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "75.5%", "reading": 82.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "82.5" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,823.5", "reading": 81.3, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "81.3" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,839.1", "reading": 75.5, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "54.0%", "reading": 75.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "75.3" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "57.5%", "reading": 75.3, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.3" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "71.8%", "reading": 73.1, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 2, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.1" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "58.1%", "reading": 65.8, "bound": null, "weight": "full", "fallback": 0.06, "sourceRank": 10, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.8" }, { "label": "回退", "display": "6%" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "52.9%", "reading": 65.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.1" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "2.9%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "anthropic/claude-sonnet-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "59.5%", "reading": 43, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 64, "sourceModelName": "claude-sonnet-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "43.0" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 63.4, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "75.0%", "reading": 105.6, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 1, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "105.6" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "32.3", "reading": 72.2, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 26, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.2" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "58.0%", "reading": 66.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "anthropic/claude-sonnet-5.5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "66.3" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "82.7%", "reading": 60, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 31, "sourceModelName": "claude-sonnet-5-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.0" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "78.0%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "claude-sonnet-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "46.5%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "claude-sonnet-5-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-29T15:15:14.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "56.8%", "reading": 43.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 55, "sourceModelName": "claude-sonnet-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "43.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 65.2, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,471", "reading": 57.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 45, "sourceModelName": "claude-sonnet-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "57.8" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-chess", "name": "Chess Puzzles" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-sonnet-5-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-sonnet-5-5", "fetched_at": "2026-10-09T04:17:00.600Z", "raw_sha256": "88371bcb6a307b485a8208bc4ce08b80765a5dd93003e28ac743c5828dea470e" } }, "claude-fable-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "releasedAt": "2026-06-09", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "inputCny": 67.023, "outputCny": 335.115, "cachedCny": 6.7023, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 6, "score": 69.9, "pendingReason": null, "onBoard": true, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 6, "score": 69.2, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 9, "score": 69.8, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 6, "score": 68.7, "coverage": 0.875, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 3, "score": 73.6, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 31, "evidence": [ { "key": "coding", "name": "编程", "score": 69.2, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 10, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "90.4%", "reading": 72.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.5" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "41.4%", "reading": 70.8, "bound": null, "weight": "full", "fallback": 0.21212121212121213, "sourceRank": 9, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.8" }, { "label": "回退", "display": "21%" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "42.4%", "reading": 69.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 25, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "55.1%", "reading": 69.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.2" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "86.0%", "reading": 68.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-fable-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.8" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "61.0%", "reading": 63.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 5, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.7" }, { "label": "回退", "display": "含回退" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 69.8, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "87.0%", "reading": 74.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "claude-fable-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-09T18:12:35.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "74.5" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "52.0%", "reading": 71.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "claude-fable-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-31T19:05:04.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "71.9" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "28.6%", "reading": 71.2, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 22, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.2" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "55.5%", "reading": 70.6, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 8, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.6" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "15.7%", "reading": 67.3, "bound": null, "weight": "full", "fallback": 0.11428571428571428, "sourceRank": 10, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.3" }, { "label": "回退", "display": "11%" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "89.7%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "claude-fable-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "41.0%", "reading": 65.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "claude-fable-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-06-09T20:20:14.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "65.2" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 68.7, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "11.3%", "reading": 73, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "73.0" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "63.6%", "reading": 71.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "71.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,539", "reading": 69.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 24, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "56.1%", "reading": 69, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.0" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "80.5%", "reading": 68.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-fable-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.9" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "48.8%", "reading": 68.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 9, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.7" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,612.7", "reading": 66.5, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 26, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.5" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "54.1%", "reading": 64, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 58, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.0" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "56.3%", "reading": 63.9, "bound": null, "weight": "full", "fallback": 0.0044444444444444444, "sourceRank": 15, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.9" }, { "label": "回退", "display": "不到 1%" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 73.6, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "64.4%", "reading": 95.1, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 4, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "95.1" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "90.7%", "reading": 85, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "claude-fable-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "85.0" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "70.7%", "reading": 81.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-fable-5_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T23:44:03.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "81.3" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "43.3", "reading": 80.7, "bound": null, "weight": "full", "fallback": "unknown", "sourceRank": 6, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "80.7" }, { "label": "回退", "display": "含回退" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.8%", "reading": 70.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "claude-fable-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.9" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "56.0%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "anthropic/claude-fable-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "82.3%", "reading": 59.4, "bound": null, "weight": "capped", "fallback": "unknown", "sourceRank": 34, "sourceModelName": "claude-fable-5", "configurationLabel": "Max 推理 · 含来源回退", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.4" }, { "label": "回退", "display": "含回退" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 70.2, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,504.3", "reading": 70.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "claude-fable-5-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "70.1" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-fable-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-fable-5", "fetched_at": "2026-10-09T04:17:00.168Z", "raw_sha256": "fc707ae84fe640f0d5f390ba3d3a12550206231abedfc20242c1061547c9e71a" } }, "gpt-6-1-sol": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "releasedAt": "2026-09-29", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "inputCny": 13.4046, "outputCny": 67.023, "cachedCny": 0.6702300000000001, "officialUrl": "https://developers.openai.com/api/docs/pricing", "note": null }, "overall": { "rank": 7, "score": 69, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 9, "score": 66.4, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 3, "score": 75.7, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 11, "score": 65.8, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 7, "score": 71.8, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 35, "evidence": [ { "key": "coding", "name": "编程", "score": 66.4, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "3.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "55.1%", "reading": 77.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 6, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "77.1" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "65.1%", "reading": 74.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "74.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "56.1%", "reading": 74.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "74.2" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "88.9%", "reading": 71.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "71.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "80.4%", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "54.2%", "reading": 54.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 58, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "78.6%", "reading": 52.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 37, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.4" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 75.7, "items": [ { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "80.0%", "reading": 87.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6.1-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-29T22:00:34.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "87.2" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "52.9%", "reading": 85.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "85.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "58.1%", "reading": 85.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "85.1" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "93.7%", "reading": 84, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6.1-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-29T22:00:34.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "84.0" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "61.0%", "reading": 79.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gpt-6.1-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-29T22:00:34.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "79.2" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "46.4%", "reading": 76.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "31.7%", "reading": 73.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "92.6%", "reading": 71.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "71.8" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "52.9%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 65.8, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "82.7%", "reading": 72.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "72.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,564.2", "reading": 70.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.7" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "50.0%", "reading": 69.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "64.9%", "reading": 69.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.3" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "60.0%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "43.0%", "reading": 66.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "66.1" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,575.1", "reading": 65, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.0" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "48.8%", "reading": 60.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.2" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "52.0%", "reading": 59.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.4" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "5.4%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 71.8, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "73.9%", "reading": 85.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "gpt-6.1-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-29T22:00:34.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "85.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "90.1%", "reading": 83.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "83.2" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "41.5", "reading": 79.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "79.3" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "33.9%", "reading": 68.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.2%", "reading": 68.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.1" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "59.0%", "reading": 67.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "openai/gpt-6.1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "67.4" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.0%", "reading": 60.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 23, "sourceModelName": "gpt-6-1-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 66.7, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,483.2", "reading": 62.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "gpt-6.1-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "62.3" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-6-1-sol", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-6-1-sol", "fetched_at": "2026-10-09T04:17:00.751Z", "raw_sha256": "4a4ae6e0836943ff861570f0b90d2e5d87a859c3fc90ee376fc546e00ca5d31a" } }, "claude-opus-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "releasedAt": "2026-07-24", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "inputCny": 33.5115, "outputCny": 167.5575, "cachedCny": 3.35115, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 8, "score": 68.9, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 7, "score": 67, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 7, "score": 70.8, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 5, "score": 69, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 6, "score": 71.9, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 36, "evidence": [ { "key": "coding", "name": "编程", "score": 67, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "3.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": 0.005, "sourceRank": 5, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" }, { "label": "回退", "display": "不到 1%" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "53.5%", "reading": 76.4, "bound": null, "weight": "full", "fallback": 0.005050505050505051, "sourceRank": 7, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.4" }, { "label": "回退", "display": "不到 1%" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "28.5%", "reading": 72.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "49.0%", "reading": 71.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.9" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "88.4%", "reading": 70.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "57.5%", "reading": 70.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.5" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "81.4%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "claude-opus-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "56.4%", "reading": 57.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.6" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 70.8, "items": [ { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "59.0%", "reading": 75.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-opus-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-25T04:37:04.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "75.3" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "27.1%", "reading": 74.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "74.2" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "85.6%", "reading": 73.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "claude-opus-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-24T19:54:20.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "73.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "28.6%", "reading": 72.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.1" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "29.1%", "reading": 71.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.5" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "37.4%", "reading": 71, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "71.0" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "54.9%", "reading": 70.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "91.2%", "reading": 68.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "claude-opus-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "42.0%", "reading": 65.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "claude-opus-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-24T18:40:39.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "65.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 69, "items": [ { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "63.6%", "reading": 78.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "78.5" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,661.6", "reading": 74.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "74.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "65.8%", "reading": 73.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "73.3" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "53.8%", "reading": 72.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.5" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,723.9", "reading": 70.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.9" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "44.0%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "58.6%", "reading": 66.3, "bound": null, "weight": "full", "fallback": 0.0007407407407407407, "sourceRank": 8, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.3" }, { "label": "回退", "display": "不到 1%" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "56.6%", "reading": 65.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.2" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "6.7%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "74.6%", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 37, "sourceModelName": "claude-opus-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 71.9, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "55.6%", "reading": 87.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 9, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "87.3" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "70.0%", "reading": 79.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "anthropic/claude-opus-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "79.5" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "88.7%", "reading": 79, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "claude-opus-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "79.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "37.1", "reading": 75.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "75.8" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "59.9%", "reading": 69, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "claude-opus-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T23:44:04.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.0" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.3%", "reading": 54.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 82, "sourceModelName": "claude-opus-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.3" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "63.8%", "reading": 52.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 26, "sourceModelName": "claude-opus-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "52.7" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 67.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,489.5", "reading": 64.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "claude-opus-5-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "64.6" } ] } ] } ], "excluded": [], "unmeasured": [], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-opus-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-opus-5", "fetched_at": "2026-10-09T04:17:00.628Z", "raw_sha256": "46550be3afcefe0c199f51fdc668070586a0c7bfd5952e51ce616e43b3c3b195" } }, "gpt-5-6-sol": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "releasedAt": "2026-07-09", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "inputCny": 26.8092, "outputCny": 134.046, "cachedCny": 2.6809200000000004, "officialUrl": "https://developers.openai.com/api/docs/pricing", "note": null }, "overall": { "rank": 9, "score": 65.9, "pendingReason": null, "onBoard": true, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 10, "score": 65.8, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 8, "score": 70.1, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 21, "score": 62.7, "coverage": 0.75, "sourceCount": 8, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 11, "score": 65.6, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 65.8, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "1.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 13, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "89.3%", "reading": 69.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.3" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "37.9%", "reading": 69.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "39.9%", "reading": 68.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.8" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "52.9%", "reading": 68.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "83.9%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-5.6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "80.5%", "reading": 65.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.2" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "20.0%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "57.1%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 70.1, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "89.1%", "reading": 76.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-5.6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-09T02:45:09.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "76.9" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "55.0%", "reading": 75, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "gpt-5.6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-09T02:07:13.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "75.0" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "58.0%", "reading": 74.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-5.6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-28T20:58:29.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "74.9" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "32.3%", "reading": 73.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "73.5" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "20.0%", "reading": 70.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.2" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "91.7%", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "gpt-5.6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "22.4%", "reading": 68.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "33.3%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "49.5%", "reading": 65.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.5" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.7, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "79.8%", "reading": 67.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "gpt-5.6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.8" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "47.5%", "reading": 67.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.8" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,478.4", "reading": 67.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.2" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "60.1%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,611.3", "reading": 66.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.8%", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "2.5%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "44.0%", "reading": 54.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "54.2" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 65.6, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "69.7%", "reading": 80.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 7, "sourceModelName": "gpt-5.6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T23:44:30.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "80.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "87.7%", "reading": 76.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-5.6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "76.3" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "22", "reading": 64.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "71.8%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-5.6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "84.0%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "26.1%", "reading": 61, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "gpt-5-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.0" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "47.0%", "reading": 55.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 71, "sourceModelName": "openai/gpt-5.6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 64.6, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,483.9", "reading": 62.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gpt-5.6-sol-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "62.6" } ] } ] } ], "excluded": [ { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1", "reason": "Pro 系统多次采样后汇总作答，不是 GPT-5.6 Sol 本身。" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor", "reason": "Pro 系统多次采样后汇总作答，不是 GPT-5.6 Sol 本身。" } ], "unmeasured": [], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-5-6-sol", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-5-6-sol", "fetched_at": "2026-10-09T04:17:01.227Z", "raw_sha256": "0b10db9dfbd93d877307df83acbaed8ac2cb3e9d95fe1280f609602dda7e05e6" } }, "gpt-6-sol": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "releasedAt": "2026-09-22", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "inputCny": 13.4046, "outputCny": 67.023, "cachedCny": 1.3404600000000002, "officialUrl": "https://developers.openai.com/api/docs/models/gpt-6-sol", "note": null }, "overall": { "rank": 10, "score": 65.7, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 8, "score": 66.5, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 10, "score": 69.3, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 17, "score": 62.9, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 14, "score": 64, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 33, "evidence": [ { "key": "coding", "name": "编程", "score": 66.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 10, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "44.4%", "reading": 72.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.2" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "57.2%", "reading": 70.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "43.9%", "reading": 70.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.2" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "87.8%", "reading": 70.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.1" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "89.3%", "reading": 69.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.3" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "81.8%", "reading": 62.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.2" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "57.6%", "reading": 59.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.2" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 69.3, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "89.8%", "reading": 77.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-22T19:59:24.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "77.8" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "30.0%", "reading": 75.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "75.5" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "56.0%", "reading": 73.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "gpt-6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-22T19:59:24.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "73.9" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "30.0%", "reading": 72.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.8" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "30.9%", "reading": 72.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.6" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "30.2%", "reading": 66.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.3" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "47.9%", "reading": 64.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.2" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "88.7%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.9, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "81.2%", "reading": 69.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "61.6%", "reading": 67.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.7" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,479.5", "reading": 67.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.3" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "43.0%", "reading": 66.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "66.1" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "54.3%", "reading": 63.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "63.8" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,507.6", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "47.1%", "reading": 58, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.0" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "49.0%", "reading": 56.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.3" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "1.7%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 64, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "85.3%", "reading": 70.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "60.7%", "reading": 69.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-6-sol_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-22T19:59:24.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "27.1", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "54.0%", "reading": 62.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "openai/gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "62.2" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.7%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "68.6%", "reading": 59.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.5" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "16.1%", "reading": 48.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 43, "sourceModelName": "gpt-6-sol", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "48.2" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 60.2, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,457.2", "reading": 52.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "gpt-6-sol-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "52.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-6-sol", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-6-sol", "fetched_at": "2026-10-09T04:17:01.923Z", "raw_sha256": "c5f8dab3a6166cd8de78755e10b124ea144b1ec438e9ea28b513208f49496f0f" } }, "muse-spark-1-3": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "releasedAt": "2026-09-02", "brand": { "src": "/model-providers/meta.svg", "monogram": "M", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "inputCny": 8.377875, "outputCny": 28.484775, "cachedCny": 1.005345, "officialUrl": "https://dev.meta.ai/docs/pricing-rate-limits/", "note": null }, "overall": { "rank": 11, "score": 65.4, "pendingReason": null, "onBoard": true, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 13, "score": 63.3, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 14, "score": 65.1, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 7, "score": 67.5, "coverage": 0.75, "sourceCount": 7, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 9, "score": 68.1, "coverage": 0.8, "sourceCount": 4, "onBoard": true } ], "metricCount": 31, "evidence": [ { "key": "coding", "name": "编程", "score": 63.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 7, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "85.9%", "reading": 68.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "33.3%", "reading": 66.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.4" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "47.4%", "reading": 65.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.2" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "20.5%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 6, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "24.7%", "reading": 62, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "81.1%", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "muse-spark-1.3-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "58.8%", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "82.1%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 65.1, "items": [ { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "36.0%", "reading": 70.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.1" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "24.9%", "reading": 68.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "89.7%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "muse-spark-1.3-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "48.7%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "74.0%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "muse-spark-1.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-18T16:34:54.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "38.0%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "muse-spark-1.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-18T19:12:25.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "10.0%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "11.0%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "25.0%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "muse-spark-1.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-18T19:12:25.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 67.5, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "23.8%", "reading": 90.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "90.2" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,583", "reading": 71.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.5" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,684.3", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "60.0%", "reading": 67.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "meta/muse_spark_1_3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.8" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "79.6%", "reading": 67.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "muse-spark-1.3-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.4" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "57.9%", "reading": 65.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.8" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "47.6%", "reading": 58.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "muse-spark-1-3-max", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "58.6" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 68.1, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "43.3%", "reading": 77.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "77.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "78.0%", "reading": 74.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "muse-spark-1.3-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "74.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "25", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "82.8%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "muse-spark-1.3-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.0%", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "muse-spark-1-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 66.2, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,494.3", "reading": 66.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "muse-spark-1.3-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "66.4" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "vals-medcode", "name": "MedCode" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/muse-spark-1-3", "api_url": "https://aihot.news/api/site/leaderboard/models/muse-spark-1-3", "fetched_at": "2026-10-09T04:17:01.916Z", "raw_sha256": "aab47bd2962f757bc78ef77263f9e37d7ab40e734b3c337b7b007e73b2e7f2a7" } }, "gemini-3-8-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "releasedAt": "2026-09-02", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "inputCny": 5.026725, "outputCny": 25.133625000000002, "cachedCny": 0.5026725, "officialUrl": "https://ai.google.dev/gemini-api/docs/pricing", "note": null }, "overall": { "rank": 12, "score": 64.5, "pendingReason": null, "onBoard": true, "coverage": 0.964, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 21, "score": 60.9, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 13, "score": 65.3, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 12, "score": 65.3, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 8, "score": 71.1, "coverage": 0.8, "sourceCount": 5, "onBoard": true } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 60.9, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "1.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 15, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "78.7%", "reading": 64.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.2" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "18.8%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "19.7%", "reading": 60.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 59, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "36.5%", "reading": 59.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.4" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "19.2%", "reading": 58.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.2" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "56.6%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "72.5%", "reading": 51.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.1" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 65.3, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "61.0%", "reading": 79.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 4, "sourceModelName": "gemini-3.8-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-02T17:15:36.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "79.2" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "36.5%", "reading": 70.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.4" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "47.0%", "reading": 69.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gemini-3.8-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-02T17:15:36.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.5" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "89.3%", "reading": 65.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "47.8%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "18.3%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "68.4%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "gemini-3.8-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-02T17:15:36.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "8.6%", "reading": 60.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.4" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "10.0%", "reading": 58.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 65.3, "items": [ { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "64.3%", "reading": 72, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gemini-3.8-flash", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "72.0" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "48.8%", "reading": 71, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "71.0" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "10.0%", "reading": 70.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "70.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "61.4%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "59.9%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,435.5", "reading": 59.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 62, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.5" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "48.1%", "reading": 59.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "google/gemini-3.8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.3" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,202.9", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 84, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "54.0%", "reading": 37.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 67, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "37.4" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 71.1, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "81.4%", "reading": 81.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "81.5" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "69.7%", "reading": 80.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "gemini-3.8-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-02T17:15:36.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "80.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "87.8%", "reading": 76.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "76.6" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "29.6", "reading": 70.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.1" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "81.3%", "reading": 57.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 46, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.6" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "21.7%", "reading": 55.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 25, "sourceModelName": "gemini-3-8-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.9" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 66, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,494.8", "reading": 66.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "gemini-3.8-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "66.6" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-3-8-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-3-8-flash", "fetched_at": "2026-10-09T04:17:01.810Z", "raw_sha256": "6373da9d8af68ba8bd0d625236648737bbdd11c3a57db26908e6f645ba19c3e5" } }, "kimi-k-3": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "releasedAt": "2026-07-16", "brand": { "src": "/model-providers/moonshot.svg", "monogram": "M", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": "https://huggingface.co/moonshotai/Kimi-K3" }, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "inputCny": 20, "outputCny": 100, "cachedCny": 2, "officialUrl": "https://platform.moonshot.cn/", "note": null }, "overall": { "rank": 13, "score": 63.3, "pendingReason": null, "onBoard": true, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 20, "score": 61.3, "coverage": 0.857, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 17, "score": 62.8, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 15, "score": 63.9, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 10, "score": 68, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 33, "evidence": [ { "key": "coding", "name": "编程", "score": 61.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 10, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "85.0%", "reading": 68, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "18.2%", "reading": 62.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.3" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "81.4%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "59.5%", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "82.1%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "17.2%", "reading": 56.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "12.6%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 75, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 62.8, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "23.4%", "reading": 67.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "90.7%", "reading": 67.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.6" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "39.0%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "kimi-k3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-16T21:59:40.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "46.9%", "reading": 63.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.3" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "72.2%", "reading": 63.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "kimi-k3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-17T01:48:42.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "63.2" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "26.0%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "kimi-k3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-29T03:26:08.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "1.4%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 63.9, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "12.9%", "reading": 76, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 8, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.0" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,500.8", "reading": 68.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.2" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "78.7%", "reading": 66.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "66.1" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "58.3%", "reading": 66, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.0" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,533.4", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "38.8%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "50.6%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "49.4%", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.1%", "reading": 60.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "kimi/kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.5" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "34.7%", "reading": 58.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "58.7" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 68, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "74.0%", "reading": 84.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 19, "sourceModelName": "moonshotai/kimi-k3", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "84.6" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "88.7%", "reading": 72.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.8" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "38.3%", "reading": 72.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "85.5%", "reading": 71.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "71.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "71.4%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "kimi-k3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "19.7", "reading": 63.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "kimi-k3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.2" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "50.6%", "reading": 59.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "kimi-k3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:36:12.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 64.1, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,488.4", "reading": 64.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "kimi-k3-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "64.2" } ] } ] } ], "excluded": [ { "key": "vals-migration", "name": "Code Migration", "reason": "Vals 确认该次运行受模型接口延迟与时限影响，不代表模型能力。" } ], "unmeasured": [ { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/kimi-k-3", "api_url": "https://aihot.news/api/site/leaderboard/models/kimi-k-3", "fetched_at": "2026-10-09T04:17:02.514Z", "raw_sha256": "6606858a983c28a6254652a5044cd44e654ac519fe50d48fd34529227a6b7b7e" } }, "claude-opus-4-8": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "releasedAt": "2026-05-28", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "inputCny": 33.5115, "outputCny": 167.5575, "cachedCny": 3.35115, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 14, "score": 63, "pendingReason": null, "onBoard": true, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 15, "score": 62.3, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 16, "score": 63.1, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 23, "score": 62.4, "coverage": 0.875, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 12, "score": 65.5, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 31, "evidence": [ { "key": "coding", "name": "编程", "score": 62.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "1.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 15, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "82.7%", "reading": 66.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "47.3%", "reading": 65.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "81.8%", "reading": 62.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "claude-opus-4-8-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "21.7%", "reading": 61.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 54, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.5" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "23.2%", "reading": 61, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "54.4%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 63.1, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "80.0%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "claude-opus-4-8_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-10T15:25:15.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "20.9%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "89.2%", "reading": 65, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "claude-opus-4-8-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.0" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "48.7%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "36.0%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "claude-opus-4-8_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-25T04:37:20.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "34.0%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "claude-opus-4-8_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-05-29T17:00:51.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "4.3%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.4, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "9.6%", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "45.0%", "reading": 65.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "53.2%", "reading": 65.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.9%", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "anthropic/claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,320.8", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 67, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,455.5", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "45.6%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 85, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "48.9%", "reading": 59.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "59.6" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "66.0%", "reading": 50, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 59, "sourceModelName": "claude-opus-4-8-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "50.0" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 65.5, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "94.0%", "reading": 127.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "anthropic/claude-opus-4.8", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "127.6" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "45.6%", "reading": 78.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 14, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "78.9" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "28.8", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "72.0%", "reading": 64.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "claude-opus-4-8-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.7" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "53.0%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "claude-opus-4-8_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:27:24.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.7%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "claude-opus-4-8-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "77.7%", "reading": 51.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 104, "sourceModelName": "claude-opus-4-8", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.8" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 62.6, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,481.5", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "claude-opus-4-8-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-opus-4-8", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-opus-4-8", "fetched_at": "2026-10-09T04:17:02.515Z", "raw_sha256": "a3be9a107d7dfdab5f7a606a208a6ba4bf09e071d84e079ec4db32b3d9354bf4" } }, "mimo-v-2-6-pro": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "releasedAt": "2026-09-21", "brand": { "src": "/model-providers/xiaomi-mimo.svg", "monogram": "X", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "inputCny": 3, "outputCny": 6, "cachedCny": 0.025, "officialUrl": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go", "note": null }, "overall": { "rank": 15, "score": 62.9, "pendingReason": null, "onBoard": true, "coverage": 0.679, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 11, "score": 64.1, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 23, "score": 60.9, "coverage": 0.571, "sourceCount": 5, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 13, "score": 65.2, "coverage": 0.75, "sourceCount": 7, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 30, "score": 59.7, "coverage": 0.4, "sourceCount": 3, "onBoard": true } ], "metricCount": 23, "evidence": [ { "key": "coding", "name": "编程", "score": 64.1, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "85.2%", "reading": 68.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "34.8%", "reading": 67, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "31.3%", "reading": 65.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.8" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "60.9%", "reading": 63.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "43.0%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 60.9, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "26.6%", "reading": 69.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "49.4%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "2.9%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "15.3%", "reading": 53.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "53.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "5.7%", "reading": 52.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 65.2, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "10.8%", "reading": 72.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.2" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,684.8", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,515.9", "reading": 68.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.8" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "59.5%", "reading": 68, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "mimo-v2.6-pro", "configurationLabel": "完整系统 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "58.6%", "reading": 66.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.2" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "57.3%", "reading": 65, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.0" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "45.0%", "reading": 55.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "xiaomi/mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 59.7, "items": [ { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "86.3%", "reading": 67.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.3" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "8.4", "reading": 55.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 75, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.5" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "18.3%", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "mimo-v2-6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 61.9, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,480.1", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "mimo-v2.6-pro", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "livebench-coding", "name": "LiveBench · 编程综合" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/mimo-v-2-6-pro", "api_url": "https://aihot.news/api/site/leaderboard/models/mimo-v-2-6-pro", "fetched_at": "2026-10-09T04:17:03.822Z", "raw_sha256": "04ca0980dda2e13556e7b6202348914e662cb870daf261a6b56be529130c6d7d" } }, "gemini-3-7-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "releasedAt": "2026-08-13", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "inputCny": 5.026725, "outputCny": 25.133625000000002, "cachedCny": 0.5026725, "officialUrl": "https://ai.google.dev/gemini-api/docs/pricing", "note": null }, "overall": { "rank": 16, "score": 62.8, "pendingReason": null, "onBoard": true, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 24, "score": 60.3, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 18, "score": 62.7, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 9, "score": 65.9, "coverage": 0.875, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 13, "score": 65.4, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 32, "evidence": [ { "key": "coding", "name": "编程", "score": 60.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "70.4%", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 37, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "57.2%", "reading": 58.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "78.9%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "34.8%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "13.6%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 72, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "12.1%", "reading": 52, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.0" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 62.7, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "47.0%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gemini-3.7-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-14T18:29:12.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "37.0%", "reading": 64.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "gemini-3.7-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-14T21:35:50.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.2" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "47.9%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "71.6%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "gemini-3.7-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-14T19:59:58.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "87.8%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "14.3%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 87, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "7.1%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 65.9, "items": [ { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "60.0%", "reading": 77.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "77.2" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "67.8%", "reading": 75, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gemini-3.7-flash", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "75.0" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "62.0%", "reading": 67.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.9" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "8.8%", "reading": 67.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.7" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "59.0%", "reading": 66.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.8" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "53.4%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,392.3", "reading": 57.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.8" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,112.5", "reading": 52.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 97, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.3" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "68.0%", "reading": 52.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 57, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "52.2" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 65.4, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "69.2%", "reading": 79.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 9, "sourceModelName": "gemini-3.7-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:30:25.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "79.5" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.9%", "reading": 78.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 3, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "78.5" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "85.5%", "reading": 70.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.9" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "26.5", "reading": 67.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.9" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "81.7%", "reading": 58.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.2" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "15.0%", "reading": 46.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 48, "sourceModelName": "gemini-3-7-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "46.5" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "31.0%", "reading": 38, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 138, "sourceModelName": "google/gemini-3.7-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "38.0" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 63.7, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,487.7", "reading": 64, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "gemini-3.7-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "64.0" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-3-7-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-3-7-flash", "fetched_at": "2026-10-09T04:17:03.008Z", "raw_sha256": "9edde67efdaf59c27086dc246732b767e9f305f7913f3377322082c33aeeea99" } }, "glm-5-3": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "releasedAt": "2026-08-18", "brand": { "src": "/model-providers/z-ai.svg", "monogram": "Z", "raster": false }, "contextWindowTokens": 1310720, "weightsUrl": "https://huggingface.co/zai-org/GLM-5.3" }, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "inputCny": 9.38322, "outputCny": 29.490120000000005, "cachedCny": 1.742598, "officialUrl": "https://docs.z.ai/guides/overview/pricing", "note": null }, "overall": { "rank": 17, "score": 62.4, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 12, "score": 63.4, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 25, "score": 60.3, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 19, "score": 62.8, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 18, "score": 62.5, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 35, "evidence": [ { "key": "coding", "name": "编程", "score": 63.4, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "1.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 13, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "38.9%", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "41.9%", "reading": 69.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.5" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "20.0%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "78.1%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.2%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "59.0%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "79.0%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "glm-5.3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 60.3, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "19.1%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "33.0%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "glm-5.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-30T23:33:33.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "68.8%", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "glm-5.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-25T14:20:20.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "23.0%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "85.8%", "reading": 59.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "glm-5.3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "42.3%", "reading": 59.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "9.5%", "reading": 58.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "5.7%", "reading": 56.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.1" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "21.0%", "reading": 47.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 80, "sourceModelName": "glm-5.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-24T19:44:18.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "47.1" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.8, "items": [ { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,508.7", "reading": 68.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.5" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,650.7", "reading": 68, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "62.2%", "reading": 68, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "8.3%", "reading": 66.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "56.6%", "reading": 65.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "65.6" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "40.5%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "55.8%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "70.2%", "reading": 54.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "glm-5.3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "54.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "42.9%", "reading": 52.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 48, "sourceModelName": "zai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.8" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 62.5, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "48.3%", "reading": 81.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 13, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "81.3" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "66.0%", "reading": 74.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 35, "sourceModelName": "z-ai/glm-5.3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "74.9" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "69.3%", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "glm-5.3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.9%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "glm-5.3", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "14.3", "reading": 59.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 70, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.5" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.7%", "reading": 54.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "glm-5-3", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.8" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "41.0%", "reading": 49.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 44, "sourceModelName": "glm-5.3_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-28T17:00:06.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "49.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 61.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,478.5", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "glm-5.3-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/glm-5-3", "api_url": "https://aihot.news/api/site/leaderboard/models/glm-5-3", "fetched_at": "2026-10-09T04:17:04.879Z", "raw_sha256": "8ea31094804fb696d514cb77efbd741f7bcd2d895816759676dec798764f7a08" } }, "gpt-5-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "releasedAt": "2026-04-23", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "inputCny": 33.5115, "outputCny": 201.06900000000002, "cachedCny": 3.35115, "officialUrl": "https://developers.openai.com/api/docs/models/gpt-5.5", "note": null }, "overall": { "rank": 18, "score": 62.3, "pendingReason": null, "onBoard": true, "coverage": 0.857, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 22, "score": 60.9, "coverage": 0.857, "sourceCount": 6, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 12, "score": 66, "coverage": 0.714, "sourceCount": 6, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 29, "score": 60.9, "coverage": 0.875, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 17, "score": 63.3, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 29, "evidence": [ { "key": "coding", "name": "编程", "score": 60.9, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "45.2%", "reading": 64, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "82.1%", "reading": 62.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.8" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "69.8%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "14.6%", "reading": 57.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 67, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.5" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "55.8%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 66, "items": [ { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "56.0%", "reading": 73.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "gpt-5.5_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-24T18:19:51.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "73.9" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "85.3%", "reading": 72.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gpt-5.5_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-11T17:47:36.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "72.8" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "27.1%", "reading": 70.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.3" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "89.7%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "45.8%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 45, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "26.0%", "reading": 52.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 62, "sourceModelName": "gpt-5.5_low", "configurationLabel": "Low 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-07T00:50:05.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "52.6" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 60.9, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "81.6%", "reading": 70.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 4, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "50.0%", "reading": 69.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "55.1%", "reading": 64.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "64.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "47.3%", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 82, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "49.1%", "reading": 60.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "51.8%", "reading": 59.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.1" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "3.8%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,353.2", "reading": 56.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 83, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.2" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,136.8", "reading": 53.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 93, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 63.3, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "87.4%", "reading": 75.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 13, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "75.5" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "63.0%", "reading": 72.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "gpt-5.5_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:28:25.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "72.4" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "20.5", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "84.3%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "70.7%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "gpt-5.5-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "47.0%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "openai/gpt-5.5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "17.8%", "reading": 50.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 38, "sourceModelName": "gpt-5-5", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 62.3, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,481.4", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "gpt-5.5-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-5-5", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-5-5", "fetched_at": "2026-10-09T04:17:04.382Z", "raw_sha256": "c08d52e9d1d159b20ba615c6b51591f17a2d7cd6e81ed762296349dc67d842a2" } }, "claude-opus-4-7": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "releasedAt": "2026-04-16", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "inputCny": 33.5115, "outputCny": 167.5575, "cachedCny": 3.35115, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 19, "score": 62.1, "pendingReason": null, "onBoard": true, "coverage": 0.643, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 16, "score": 62.3, "coverage": 0.571, "sourceCount": 4, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 24, "score": 60.5, "coverage": 0.429, "sourceCount": 4, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 22, "score": 62.6, "coverage": 0.625, "sourceCount": 6, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 21, "score": 62.2, "coverage": 1, "sourceCount": 4, "onBoard": true } ], "metricCount": 20, "evidence": [ { "key": "coding", "name": "编程", "score": 62.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "43.9%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "82.1%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "claude-opus-4-7-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "71.0%", "reading": 60.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 60.5, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "70.2%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "claude-opus-4-7_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-10T21:41:54.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "87.2%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "claude-opus-4-7-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "28.0%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "claude-opus-4-7_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-25T01:43:24.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "7.0%", "reading": 24.9, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 148, "sourceModelName": "claude-opus-4-7_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-06T22:19:09.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "24.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.6, "items": [ { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "54.9%", "reading": 67.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.5" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "78.3%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "claude-opus-4-7-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "43.8%", "reading": 65, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.0" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "6.7%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "49.2%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "claude-opus-4-7", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "51.5%", "reading": 58.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "anthropic/claude-opus-4-7", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.8" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 62.2, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "74.0%", "reading": 84.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 28, "sourceModelName": "anthropic/claude-opus-4.7", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "84.6" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "28.9%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "claude-opus-4-7-non-reasoning", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "51.7%", "reading": 60.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "claude-opus-4-7_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:27:26.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "60.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "77.9%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "claude-opus-4-7-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "66.7%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "claude-opus-4-7-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 64.9, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,501.3", "reading": 69, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-opus-4-7-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "69.0" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0 · AA" }, { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "artificial-analysis-scicode", "name": "SciCode · AA" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "artificial-analysis-hle", "name": "HLE · AA" }, { "key": "artificial-analysis-critpt", "name": "CritPt · AA" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "artificial-analysis-gdpval", "name": "GDPval-AA" }, { "key": "artificial-analysis-briefcase", "name": "AA-Briefcase" }, { "key": "artificial-analysis-automationbench", "name": "AutomationBench · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "artificial-analysis-omniscience", "name": "AA-Omniscience" }, { "key": "artificial-analysis-lcr", "name": "AA-LCR" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-opus-4-7", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-opus-4-7", "fetched_at": "2026-10-09T04:17:05.670Z", "raw_sha256": "857b623b8d151b85abd4c95f04dcd1933ef74e22c3796a66fe4a8a593ed40f8c" } }, "gpt-5-6-terra": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "releasedAt": "2026-07-09", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "inputCny": 13.4046, "outputCny": 80.4276, "cachedCny": 1.3404600000000002, "officialUrl": "https://developers.openai.com/api/docs/models", "note": null }, "overall": { "rank": 20, "score": 62.1, "pendingReason": null, "onBoard": true, "coverage": 0.964, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 18, "score": 61.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 11, "score": 66.1, "coverage": 0.857, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 27, "score": 61.6, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 27, "score": 59.8, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 61.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "35.4%", "reading": 67.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.2" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "47.8%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "74.6%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "22.7%", "reading": 60.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.7" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "14.8%", "reading": 58, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "78.2%", "reading": 57.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.6" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "55.0%", "reading": 55.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.8" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 66.1, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "54.0%", "reading": 74.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gpt-5.6-terra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-09T17:07:54.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "74.3" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "86.0%", "reading": 73.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "gpt-5.6-terra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-09T15:58:30.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "73.5" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "30.0%", "reading": 72.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "90.6%", "reading": 67.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.5" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "35.0%", "reading": 63, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "gpt-5.6-terra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-28T20:58:46.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "63.0" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "13.8%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "10.0%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "42.9%", "reading": 59.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 56, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 61.6, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "79.3%", "reading": 67, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.0" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "58.2%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "59.6%", "reading": 66.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.7" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "38.8%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "54.4%", "reading": 61.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.9" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,333.5", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,453.2", "reading": 60.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 59, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.2" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "0.8%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "43.4%", "reading": 53.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "53.5" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 59.8, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "31.7%", "reading": 66.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "82.9%", "reading": 65.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.6" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "54.0%", "reading": 62.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 62, "sourceModelName": "openai/gpt-5.6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "62.2" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.0%", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "64.6%", "reading": 53.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "gpt-5.6-terra-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "53.9" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "43.2%", "reading": 51.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "gpt-5.6-terra_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T23:44:21.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "51.6" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "0.1", "reading": 49.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 98, "sourceModelName": "gpt-5-6-terra", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "49.9" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 59.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,465.5", "reading": 55.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "gpt-5.6-terra-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "55.8" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-5-6-terra", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-5-6-terra", "fetched_at": "2026-10-09T04:17:04.385Z", "raw_sha256": "23473759aa0a4cbb50ab6faf9c38be3ed71d73a3bb17d8604a77dac12ea3b188" } }, "grok-4-7": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "releasedAt": "2026-09-21", "brand": { "src": "/model-providers/xai.svg", "monogram": "X", "raster": false }, "contextWindowTokens": 500000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "inputCny": 13.4046, "outputCny": 40.2138, "cachedCny": 3.35115, "officialUrl": "https://docs.x.ai/developers/grok-4-7", "note": null }, "overall": { "rank": 21, "score": 61.6, "pendingReason": null, "onBoard": true, "coverage": 0.964, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 17, "score": 61.7, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 26, "score": 60, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 14, "score": 64.4, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 22, "score": 62.1, "coverage": 0.8, "sourceCount": 5, "onBoard": true } ], "metricCount": 32, "evidence": [ { "key": "coding", "name": "编程", "score": 61.7, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "86.2%", "reading": 68.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.8" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "28.8%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.8%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "25.8%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "83.9%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "57.4%", "reading": 58.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.9" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.2%", "reading": 56.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 60, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "17.7%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "38.0%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "grok-4.7_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-22T21:49:22.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "25.2%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "10.0%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "43.1%", "reading": 60.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.1" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "29.0%", "reading": 59.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "grok-4.7_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-22T21:49:22.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "82.7%", "reading": 56.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.1" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "53.0%", "reading": 53.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "grok-4.7_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-22T21:49:22.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "53.3" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 64.4, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "12.5%", "reading": 75.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 10, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "75.3" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,643.8", "reading": 74, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "74.0" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,712.5", "reading": 70.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 8, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "65.6%", "reading": 69.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "54.6%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "grok-4.7", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "76.9%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "49.6%", "reading": 61, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.0" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "52.3%", "reading": 59.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "grok/grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.6" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "32.6%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "grok-4.7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 62.1, "items": [ { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "32", "reading": 72, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.3%", "reading": 70, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.0" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "56.0%", "reading": 64.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "grok-4.7_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-22T21:49:22.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.9" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "80.1%", "reading": 60.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.5" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "76.7%", "reading": 50.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 115, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.3" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "15.0%", "reading": 46.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 48, "sourceModelName": "grok-4-7", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "46.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 55.8, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,442.5", "reading": 47.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 91, "sourceModelName": "grok-4.7-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "47.2" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/grok-4-7", "api_url": "https://aihot.news/api/site/leaderboard/models/grok-4-7", "fetched_at": "2026-10-09T04:17:06.780Z", "raw_sha256": "0d1afcfa3e64e40b26ceaeb673612f544cff0d202aac08362928bf868427e1c3" } }, "grok-4-6": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "releasedAt": "2026-08-12", "brand": { "src": "/model-providers/xai.svg", "monogram": "X", "raster": false }, "contextWindowTokens": 500000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "inputCny": 13.4046, "outputCny": 40.2138, "cachedCny": 3.35115, "officialUrl": "https://docs.x.ai/developers/grok-4-6", "note": null }, "overall": { "rank": 22, "score": 61.3, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 30, "score": 58.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 20, "score": 62.3, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 10, "score": 65.8, "coverage": 1, "sourceCount": 10, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 19, "score": 62.5, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 36, "evidence": [ { "key": "coding", "name": "编程", "score": 58.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "1.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 15, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.6%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "76.2%", "reading": 62.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.9" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "17.2%", "reading": 59.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 63, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.1" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "14.8%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "17.2%", "reading": 56.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.7" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "76.8%", "reading": 55.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "grok-4.6", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.9" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "80.4%", "reading": 54.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "54.6" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "53.0%", "reading": 53.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 62.3, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "90.5%", "reading": 67.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "grok-4.6", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.3" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "30.6%", "reading": 66.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.6" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "19.7%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "11.4%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "34.0%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "grok-4.6_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-14T20:27:46.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "44.1%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 50, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "66.0%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "grok-4.6_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-14T15:46:50.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "31.0%", "reading": 57.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "grok-4.6_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-14T13:47:42.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "57.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "6.2%", "reading": 53.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "grok-4-6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.2" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 65.8, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "15.8%", "reading": 80.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 6, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "80.6" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "65.3%", "reading": 72.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "72.8" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "67.0%", "reading": 70.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,544.4", "reading": 69.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.9" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "45.7%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,646.7", "reading": 67.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.9" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "41.3%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "grok-4-6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.7%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "73.9%", "reading": 59.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "grok-4.6", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.4" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "44.7%", "reading": 55.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 39, "sourceModelName": "grok/grok-4.6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 62.5, "items": [ { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "29.3", "reading": 70, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "70.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "83.7%", "reading": 67.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "grok-4.6", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.2" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "71.9%", "reading": 64.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "grok-4.6", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.5" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "56.0%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "x-ai/grok-4.6", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "48.9%", "reading": 57.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "grok-4.6_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:29:19.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "57.5" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "81.0%", "reading": 57, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "grok-4-6-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.0" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "12.2%", "reading": 41.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 57, "sourceModelName": "grok-4-6", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "41.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 57.7, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,453.8", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "grok-4.6-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] } ] } ], "excluded": [], "unmeasured": [], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/grok-4-6", "api_url": "https://aihot.news/api/site/leaderboard/models/grok-4-6", "fetched_at": "2026-10-09T04:17:06.231Z", "raw_sha256": "d2cb8899467f3f68e0365eb5fc64f93e35fb7bfadf90720e4c659c128206fd07" } }, "muse-spark-1-2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "releasedAt": "2026-08-05", "brand": { "src": "/model-providers/meta.svg", "monogram": "M", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "inputCny": 8.377875, "outputCny": 28.484775, "cachedCny": 1.005345, "officialUrl": "https://dev.meta.ai/docs/pricing-rate-limits/", "note": null }, "overall": { "rank": 23, "score": 61.1, "pendingReason": null, "onBoard": true, "coverage": 0.821, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 28, "score": 58.9, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 15, "score": 63.2, "coverage": 0.429, "sourceCount": 3, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 25, "score": 62.1, "coverage": 0.875, "sourceCount": 8, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 20, "score": 62.3, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 27, "evidence": [ { "key": "coding", "name": "编程", "score": 58.9, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "79.1%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "57.4%", "reading": 58.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.9" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.5%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "muse-spark-1.2-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "30.0%", "reading": 55.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "7.1%", "reading": 50.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 89, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "6.1%", "reading": 43.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 37, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "43.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 63.2, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "90.0%", "reading": 66.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "muse-spark-1.2-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "66.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "17.7%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "45.5%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.1, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "25.4%", "reading": 91.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 1, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "91.9" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "60.6%", "reading": 68.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.5" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "76.5%", "reading": 62.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "muse-spark-1.2-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.8" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,477.5", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,329.1", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 66, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "49.3%", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "meta/muse_spark_1_2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "40.6%", "reading": 57.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 91, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.6" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "36.4%", "reading": 49.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 47, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "49.5" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 62.3, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "60.3%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "muse-spark-1.2_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:36:49.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "27.2", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.3%", "reading": 68.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "muse-spark-1.2-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.4" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "31.1%", "reading": 66.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "78.6%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "muse-spark-1.2-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "49.0%", "reading": 57.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 85, "sourceModelName": "meta/muse-spark-1.2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "57.2" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.0%", "reading": 53.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 90, "sourceModelName": "muse-spark-1-2", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.8" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 63.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,493.5", "reading": 66.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "muse-spark-1.2 (xHigh)", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "66.1" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/muse-spark-1-2", "api_url": "https://aihot.news/api/site/leaderboard/models/muse-spark-1-2", "fetched_at": "2026-10-09T04:17:06.194Z", "raw_sha256": "d6ef052aef16eda24842cd745602722ad27a08daaa35cf33437aa15606f5c803" } }, "step-5-preview": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "releasedAt": "2026-09-18", "brand": { "src": "/model-providers/stepfun.svg", "monogram": "S", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": null, "overall": { "rank": 24, "score": 60.6, "pendingReason": null, "onBoard": true, "coverage": 0.571, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 19, "score": 61.4, "coverage": 0.571, "sourceCount": 5, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 30, "score": 59.2, "coverage": 0.571, "sourceCount": 5, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 26, "score": 61.8, "coverage": 0.625, "sourceCount": 6, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 26, "score": 60.2, "coverage": 0.4, "sourceCount": 3, "onBoard": true } ], "metricCount": 20, "evidence": [ { "key": "coding", "name": "编程", "score": 61.4, "items": [ { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "33.3%", "reading": 66.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.4" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "31.8%", "reading": 66, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "58.9%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "39.0%", "reading": 60.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "70.4%", "reading": 60.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 59.2, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "20.9%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "46.5%", "reading": 63, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.0" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "18.5%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "2.9%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "2.4%", "reading": 50.9, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数 · 至多", "display": "50.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 61.8, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "9.2%", "reading": 68.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,584.2", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,425.2", "reading": 65.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.1" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "51.0%", "reading": 62.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 72, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.6" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "35.0%", "reading": 58.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.1" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "50.7%", "reading": 58, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "stepfun/step-5-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.0" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 60.2, "items": [ { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "88.3%", "reading": 72, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "72.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "16.4", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 69, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "16.7%", "reading": 49.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 42, "sourceModelName": "step-5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "49.1" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 56.8, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,453.7", "reading": 51.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "Step 5 Preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "51.4" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "livebench-coding", "name": "LiveBench · 编程综合" }, { "key": "vals-program", "name": "ProgramBench" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析" }, { "key": "vals-medcode", "name": "MedCode" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/step-5-preview", "api_url": "https://aihot.news/api/site/leaderboard/models/step-5-preview", "fetched_at": "2026-10-09T04:17:06.255Z", "raw_sha256": "eca0fef9870e63d2a1f55902c770d9a6763e185450d011b3e91d6c9d2b8a6ab2" } }, "claude-haiku-5-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-haiku-5-5", "name": "Claude Haiku 5.5", "provider": "Anthropic", "releasedAt": "2026-10-07", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": null, "overall": { "rank": 25, "score": 60.6, "pendingReason": null, "onBoard": true, "coverage": 0.821, "missingDimensions": [ "真人盲选" ] }, "categories": [ { "key": "coding", "name": "编程", "rank": 14, "score": 62.7, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 19, "score": 62.6, "coverage": 0.714, "sourceCount": 6, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 34, "score": 59.5, "coverage": 0.875, "sourceCount": 8, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 38, "score": 54.1, "coverage": 0.8, "sourceCount": 4, "onBoard": false } ], "metricCount": 28, "evidence": [ { "key": "coding", "name": "编程", "score": 62.7, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "2.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 7, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "90.4%", "reading": 72.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 3, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.5" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "89.3%", "reading": 69.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 1, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.3" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "35.4%", "reading": 67.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.8" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "32.8%", "reading": 66.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.2" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "20.4%", "reading": 64.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 7, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.7" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.6%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.9%", "reading": 57.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "claude-haiku-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.2" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "55.0%", "reading": 55.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.8" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 62.6, "items": [ { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "35.1%", "reading": 69.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 10, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "69.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "20.0%", "reading": 67.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.2" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "18.9%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "10.0%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "44.4%", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "84.6%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "claude-haiku-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 59.5, "items": [ { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,577.1", "reading": 71.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 18, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,618.1", "reading": 66.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "56.1%", "reading": 65.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "65.2" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "51.5%", "reading": 63.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.5" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "54.1%", "reading": 61.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.6" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "1.3%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "anthropic/claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "35.4%", "reading": 55, "bound": null, "weight": "full", "fallback": null, "sourceRank": 102, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.0" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "51.8%", "reading": 35.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 69, "sourceModelName": "claude-haiku-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "35.2" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 54.1, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "63.3%", "reading": 94.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 5, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "94.1" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "82.7%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "10.7", "reading": 57, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "claude-haiku-5-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "54.7%", "reading": 40.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 67, "sourceModelName": "claude-haiku-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "40.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "62.6%", "reading": 36.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 67, "sourceModelName": "claude-haiku-5-5-max-effort", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "36.1" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" }, { "key": "arena-text", "name": "Arena Text" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-haiku-5-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-haiku-5-5", "fetched_at": "2026-10-09T04:17:07.859Z", "raw_sha256": "07fa3216ec32d9d4e87db6bc0e4e95abff1f60f597dde3f36fb108e0da0cad48" } }, "deepseek-v-4-1-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "deepseek-v-4-1-flash", "name": "DeepSeek V4.1 Flash", "provider": "DeepSeek", "releasedAt": "2026-09-10", "brand": { "src": "/model-providers/deepseek.svg", "monogram": "D", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash" }, "price": { "currency": "USD", "input": 0.3, "output": 1.2, "cached": 0.006, "inputCny": 2.01069, "outputCny": 8.04276, "cachedCny": 0.0402138, "officialUrl": "https://api-docs.deepseek.com/quick_start/pricing/", "note": null }, "overall": { "rank": 26, "score": 60.1, "pendingReason": null, "onBoard": true, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 23, "score": 60.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 32, "score": 58.7, "coverage": 0.714, "sourceCount": 6, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 28, "score": 61.1, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 28, "score": 59.8, "coverage": 1, "sourceCount": 5, "onBoard": true } ], "metricCount": 31, "evidence": [ { "key": "coding", "name": "编程", "score": 60.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "84.7%", "reading": 67.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.8" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "45.6%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "26.8%", "reading": 63.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 45, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "16.4%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "80.0%", "reading": 59.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 2, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.9" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "19.7%", "reading": 58.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.6" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "51.9%", "reading": 51.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 82, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.8" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 58.7, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "86.7%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "14.3%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 87, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "21.2%", "reading": 59.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.4" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "9.0%", "reading": 57.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.5" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "39.2%", "reading": 56.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 80, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.7" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "0.0%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 61.1, "items": [ { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "68.9%", "reading": 71.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 4, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.5" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "79.3%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,600", "reading": 66, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.0" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,420.5", "reading": 64.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.9" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "6.7%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "37.5%", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.5%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "39.5%", "reading": 52.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "52.1" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "41.2%", "reading": 50.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 56, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 59.8, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "57.0%", "reading": 65.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "deepseek/deepseek-v4.1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "65.3" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "84.0%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "81.2%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "70.0%", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "22.8%", "reading": 57.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.2" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "-5.3", "reading": 46.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 113, "sourceModelName": "deepseek-v4-1-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "46.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 59.6, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,474.4", "reading": 59.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "deepseek-v4.1-flash-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "59.1" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/deepseek-v-4-1-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/deepseek-v-4-1-flash", "fetched_at": "2026-10-09T04:17:07.348Z", "raw_sha256": "c0d42dc40cfb9b7565d6622dee57a506bec3a933321cd3b7a1f341cff78fb91e" } }, "deepseek-v-4-pro-20260813": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "deepseek-v-4-pro-20260813", "name": "DeepSeek V4 Pro 0813", "provider": "DeepSeek", "releasedAt": "2026-08-13", "brand": { "src": "/model-providers/deepseek.svg", "monogram": "D", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813" }, "price": { "currency": "CNY", "input": 9, "output": 27, "cached": 0.3, "inputCny": 9, "outputCny": 27, "cachedCny": 0.3, "officialUrl": "https://api-docs.deepseek.com/zh-cn/quick_start/pricing/", "note": null }, "overall": { "rank": 27, "score": 58.8, "pendingReason": null, "onBoard": true, "coverage": 0.964, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 29, "score": 58.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 27, "score": 59.8, "coverage": 0.857, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 31, "score": 60.5, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 34, "score": 56.3, "coverage": 1, "sourceCount": 6, "onBoard": false } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 58.5, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "82.3%", "reading": 66.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "66.2" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "41.5%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "17.5%", "reading": 61.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "14.1%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 70, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.2%", "reading": 56.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "deepseek-v4-pro-0813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.3" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "14.1%", "reading": 54, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "54.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "51.0%", "reading": 50.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 92, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.8" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 59.8, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "47.0%", "reading": 69.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "deepseek-v4-pro-0813_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-18T21:03:23.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "69.4" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "43.0%", "reading": 67.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "deepseek-v4-pro-0813_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-19T18:14:49.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "67.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "18.0%", "reading": 63.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 62, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.3" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "85.8%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "deepseek-v4-pro-0813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "64.6%", "reading": 59, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "deepseek-v4-pro-0813_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-19T13:35:48.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.0" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "41.0%", "reading": 58.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 71, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.3" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "4.3%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "5.7%", "reading": 52.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 60.5, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "79.2%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "deepseek-v4-pro-0813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "56.7%", "reading": 65.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.3" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "7.5%", "reading": 64.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.5" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "40.2%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "deepseek-v4-pro-08-13", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,454.1", "reading": 60.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.2" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "47.3%", "reading": 58.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "deepseek-v4-pro-08-13", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "58.3" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,256.4", "reading": 58.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 79, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.2" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "50.4%", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "42.5%", "reading": 52.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 56.3, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "82.1%", "reading": 64, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "deepseek-v4-pro-0813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.0" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "52.9%", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "deepseek-v4-pro-0813_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:30:20.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "67.7%", "reading": 58.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "deepseek-v4-pro-0813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.2" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "80.3%", "reading": 55.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.9" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "17.8%", "reading": 50.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "0.8", "reading": 50.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 95, "sourceModelName": "deepseek-v4-pro", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.4" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "35.0%", "reading": 42.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 114, "sourceModelName": "deepseek/deepseek-v4-pro-0813", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "42.6" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 57.3, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,464.3", "reading": 55.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 56, "sourceModelName": "deepseek-v4-pro-high-20260813", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "55.3" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/deepseek-v-4-pro-20260813", "api_url": "https://aihot.news/api/site/leaderboard/models/deepseek-v-4-pro-20260813", "fetched_at": "2026-10-09T04:17:07.968Z", "raw_sha256": "df6ef68a64298a719302ed27f697000a83e5c20f22a2021cd405ecf54f68da0a" } }, "claude-sonnet-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "claude-sonnet-5", "name": "Claude Sonnet 5", "provider": "Anthropic", "releasedAt": "2026-06-30", "brand": { "src": "/model-providers/anthropic.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "inputCny": 13.4046, "outputCny": 67.023, "cachedCny": 1.3404600000000002, "officialUrl": "https://platform.claude.com/docs/en/about-claude/pricing", "note": null }, "overall": { "rank": 28, "score": 58.8, "pendingReason": null, "onBoard": true, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 31, "score": 58.4, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 31, "score": 59.2, "coverage": 0.857, "sourceCount": 7, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 33, "score": 59.6, "coverage": 0.875, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 31, "score": 58.5, "coverage": 1, "sourceCount": 6, "onBoard": false } ], "metricCount": 33, "evidence": [ { "key": "coding", "name": "编程", "score": 58.4, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "81.3%", "reading": 65.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.6" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.4%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "80.7%", "reading": 60.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "claude-sonnet-5-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "60.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "14.1%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 70, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "82.1%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "13.8%", "reading": 56.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.5" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "54.3%", "reading": 54.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 56, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.9" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "9.6%", "reading": 49, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "49.0" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 59.2, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "88.7%", "reading": 64.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "claude-sonnet-5-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "64.1" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "35.0%", "reading": 63, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "claude-sonnet-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-28T21:09:50.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "63.0" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "16.9%", "reading": 62.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.3" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "65.6%", "reading": 59.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "claude-sonnet-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-30T21:54:08.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.6" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "41.3%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 69, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "5.7%", "reading": 56.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.1" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "16.0%", "reading": 39.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 109, "sourceModelName": "claude-sonnet-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-06-30T18:32:59.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "39.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 59.6, "items": [ { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "46.3%", "reading": 66.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.9" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "54.5%", "reading": 64, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "claude-sonnet-5-max", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "64.0" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,358", "reading": 62.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.3" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "53.9%", "reading": 61.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,465.1", "reading": 60.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 54, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.7" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "47.5%", "reading": 58.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "58.6" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "71.7%", "reading": 56.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "claude-sonnet-5-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.7" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "5.0%", "reading": 56.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "36.5%", "reading": 55.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 97, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.6" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 58.5, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "78.0%", "reading": 90.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 14, "sourceModelName": "anthropic/claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "90.1" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "55.0%", "reading": 86.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 10, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "86.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "16.5", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 68, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "82.0%", "reading": 58.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "claude-sonnet-5", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "63.9%", "reading": 52.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "claude-sonnet-5-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "52.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.0%", "reading": 52.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "claude-sonnet-5-xhigh-effort", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "52.2" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "33.7%", "reading": 41.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 56, "sourceModelName": "claude-sonnet-5_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:27:42.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "41.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 57, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,461.8", "reading": 54.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "claude-sonnet-5-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "54.4" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/claude-sonnet-5", "api_url": "https://aihot.news/api/site/leaderboard/models/claude-sonnet-5", "fetched_at": "2026-10-09T04:17:07.883Z", "raw_sha256": "3d67c7f06974211c6b06bf78dfc76b500481093977421f158ebc88eddab89332" } }, "qwen-3-8-max": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "qwen-3-8-max", "name": "Qwen3.8 Max", "provider": "Alibaba", "releasedAt": "2026-07-19", "brand": { "src": "/model-providers/qianwen.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 1.5, "inputCny": 12, "outputCny": 36, "cachedCny": 1.5, "officialUrl": "https://help.aliyun.com/zh/model-studio/qwen3-8-max", "note": null }, "overall": { "rank": 29, "score": 58.6, "pendingReason": null, "onBoard": true, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 38, "score": 54.7, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 21, "score": 62.2, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 16, "score": 63.5, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 33, "score": 56.8, "coverage": 1, "sourceCount": 6, "onBoard": false } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 54.7, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "34.3%", "reading": 67.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 13, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "18.7%", "reading": 59.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.9" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "64.7%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "12.8%", "reading": 55, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "53.2%", "reading": 53.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 70, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "72.9%", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "24.0%", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "57.1%", "reading": 31.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 44, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "31.5" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 62.2, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "20.0%", "reading": 65, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.0" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "12.9%", "reading": 64.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.9" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "74.7%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "qwen3.8-max_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-04T19:32:51.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "38.0%", "reading": 64.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "qwen3.8-max_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-05T01:42:38.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "88.2%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "23.9%", "reading": 61.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.6" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "43.0%", "reading": 60, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.0" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "29.0%", "reading": 55.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 54, "sourceModelName": "qwen3.8-max_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-04T16:28:46.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "55.5" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 63.5, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "10.4%", "reading": 71.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "71.4" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "63.3%", "reading": 71.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "qwen-3-8-max-xhigh", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "71.1" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "46.7%", "reading": 69.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "qwen-3-8-max-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "69.2" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,607.7", "reading": 66.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.3" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "78.4%", "reading": 65.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.6" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,386.5", "reading": 63.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "49.2%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 76, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "50.6%", "reading": 57.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "40.7%", "reading": 50.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 61, "sourceModelName": "alibaba/qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "50.1" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 56.8, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.1%", "reading": 68, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 22, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.7%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "45.8%", "reading": 54.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 37, "sourceModelName": "qwen3.8-max_xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:30:55.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "54.3" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "19.4%", "reading": 53, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.0" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "78.3%", "reading": 52.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 95, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "3.4", "reading": 52.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 86, "sourceModelName": "qwen3-8-max-0803", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.1" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "27.0%", "reading": 33.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 118, "sourceModelName": "qwen/qwen3.8-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "33.1" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 60.3, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,481.6", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "qwen3.8-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/qwen-3-8-max", "api_url": "https://aihot.news/api/site/leaderboard/models/qwen-3-8-max", "fetched_at": "2026-10-09T04:17:08.413Z", "raw_sha256": "c71c90cfb466537854cb446928fee5c2a0b5d07d9b94a70ebd45069aa1aceaba" } }, "muse-spark-1-1": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "muse-spark-1-1", "name": "Muse Spark 1.1", "provider": "Meta", "releasedAt": "2026-07-09", "brand": { "src": "/model-providers/meta.svg", "monogram": "M", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "inputCny": 8.377875, "outputCny": 28.484775, "cachedCny": 1.005345, "officialUrl": "https://dev.meta.ai/docs/pricing-rate-limits/", "note": null }, "overall": { "rank": 30, "score": 58.5, "pendingReason": null, "onBoard": true, "coverage": 0.714, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 33, "score": 57.3, "coverage": 0.857, "sourceCount": 6, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 22, "score": 60.9, "coverage": 0.429, "sourceCount": 3, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 38, "score": 57.7, "coverage": 0.75, "sourceCount": 7, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 32, "score": 58.3, "coverage": 0.8, "sourceCount": 4, "onBoard": false } ], "metricCount": 22, "evidence": [ { "key": "coding", "name": "编程", "score": 57.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "meta/muse_spark_1_1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "72.2%", "reading": 61, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "meta/muse_spark_1_1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "58.8%", "reading": 60.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 15, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.8" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.2%", "reading": 56.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "muse-spark-1.1-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.3" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "31.1%", "reading": 56.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 37, "sourceModelName": "meta/muse_spark_1_1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.2" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "6.1%", "reading": 49.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 93, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "49.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 60.9, "items": [ { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "46.2%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "87.7%", "reading": 62.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "muse-spark-1.1-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.6" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "15.1%", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 83, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 57.7, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "20.0%", "reading": 86, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 4, "sourceModelName": "meta/muse_spark_1_1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "86.0" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "57.2%", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "meta/muse_spark_1_1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "72.5%", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "muse-spark-1.1-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "38.8%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 95, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,214.8", "reading": 50.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 105, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "31.8%", "reading": 45.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 50, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "45.5" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "847.6", "reading": 41.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 136, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "41.5" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 58.3, "items": [ { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "28.1", "reading": 69.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 38, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "69.1" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "57.8%", "reading": 66.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "muse-spark-1.1", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-31T19:21:58.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "66.7" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "69.6%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "muse-spark-1.1-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "77.7%", "reading": 51.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 104, "sourceModelName": "muse-spark-1-1", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.3%", "reading": 51.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "muse-spark-1.1-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 61.4, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,491.3", "reading": 65.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "muse-spark-1.1", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "65.3" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "vals-medcode", "name": "MedCode" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/muse-spark-1-1", "api_url": "https://aihot.news/api/site/leaderboard/models/muse-spark-1-1", "fetched_at": "2026-10-09T04:17:08.541Z", "raw_sha256": "6f39f82fc15b1b93e43534d404debad6a17f15767ae9f7c475452f97d6b42a59" } }, "gpt-6-luna": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-6-luna", "name": "GPT-6 Luna", "provider": "OpenAI", "releasedAt": "2026-09-22", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 0.1, "output": 0.5, "cached": 0.01, "inputCny": 0.6702300000000001, "outputCny": 3.35115, "cachedCny": 0.067023, "officialUrl": "https://developers.openai.com/api/docs/models/gpt-6-luna", "note": null }, "overall": { "rank": 33, "score": 57.7, "pendingReason": null, "onBoard": false, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 25, "score": 59.6, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 36, "score": 57.9, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 36, "score": 58, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 39, "score": 53.5, "coverage": 1, "sourceCount": 6, "onBoard": false } ], "metricCount": 34, "evidence": [ { "key": "coding", "name": "编程", "score": 59.6, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "81.6%", "reading": 65.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "87.5%", "reading": 65.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.7" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "42.6%", "reading": 62.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 25, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.7" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "79.0%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "12.6%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 75, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "54.6%", "reading": 55.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "13.6%", "reading": 53.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "53.6" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 57.9, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "78.9%", "reading": 67.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "gpt-6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-09-22T22:14:05.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "67.6" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "19.4%", "reading": 64.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 56, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.6" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "19.4%", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "31.0%", "reading": 57.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "gpt-6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-09-22T22:14:05.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "57.3" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "8.6%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "38.5%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 85, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "81.8%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "4.3%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "7.0%", "reading": 40.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 117, "sourceModelName": "gpt-6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-09-22T22:14:05.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "40.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 58, "items": [ { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "40.2%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 13, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "53.2%", "reading": 63.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.6" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,336.4", "reading": 61.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 63, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.5" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,431.6", "reading": 59.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.3" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "73.4%", "reading": 58.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.7" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "49.9%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "2.9%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "44.3%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "44.7%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 53.5, "items": [ { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "58.0%", "reading": 66.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 52, "sourceModelName": "openai/gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "66.3" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.3%", "reading": 61.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.2" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "73.8%", "reading": 50.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "50.6" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "0.7", "reading": 50.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 97, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.3" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "41.4%", "reading": 49.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "gpt-6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-09-22T22:14:05.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "49.7" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "16.1%", "reading": 48.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "gpt-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "48.2" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "55.9%", "reading": 42.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 64, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "42.4" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 53.1, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,442.9", "reading": 47.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 90, "sourceModelName": "gpt-6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "47.4" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-6-luna", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-6-luna", "fetched_at": "2026-10-09T04:17:09.232Z", "raw_sha256": "e8accda6ccc17f958d815b6f8ccffdbef0a89d9d9262277c223b0de58461ad0d" } }, "mimo-v-2-6-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "mimo-v-2-6-flash", "name": "MiMo V 2.6 Flash", "provider": "Xiaomi", "releasedAt": "2026-09-21", "brand": { "src": "/model-providers/xiaomi-mimo.svg", "monogram": "X", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": null, "overall": { "rank": 35, "score": 57.6, "pendingReason": null, "onBoard": false, "coverage": 0.643, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 26, "score": 59.3, "coverage": 0.857, "sourceCount": 7, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 40, "score": 55.5, "coverage": 0.571, "sourceCount": 4, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 18, "score": 62.8, "coverage": 0.625, "sourceCount": 6, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 50, "score": 49.8, "coverage": 0.4, "sourceCount": 2, "onBoard": false } ], "metricCount": 20, "evidence": [ { "key": "coding", "name": "编程", "score": 59.3, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.5%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 18, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "79.0%", "reading": 64.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "64.3" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "22.7%", "reading": 62, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.0" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "40.9%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 27, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "24.2%", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "51.3%", "reading": 51.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 90, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.1" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 55.5, "items": [ { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "21.6%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "12.0%", "reading": 57.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 96, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "4.3%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "35.1%", "reading": 52.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 98, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "52.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.8, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "11.3%", "reading": 73, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 11, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "73.0" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "64.1%", "reading": 68.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "68.9" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,494.7", "reading": 67.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.9" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,610.5", "reading": 66.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.4" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "56.3%", "reading": 63.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.8" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "41.1%", "reading": 50.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 57, "sourceModelName": "xiaomi/mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "50.6" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 49.8, "items": [ { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "74.3%", "reading": 47, "bound": null, "weight": "full", "fallback": null, "sourceRank": 127, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "47.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "-12.7", "reading": 41.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 136, "sourceModelName": "mimo-v2-6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "41.4" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 54.3, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,451.8", "reading": 50.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 77, "sourceModelName": "mimo-v2.6-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "livebench-coding", "name": "LiveBench · 编程综合" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/mimo-v-2-6-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/mimo-v-2-6-flash", "fetched_at": "2026-10-09T04:17:08.551Z", "raw_sha256": "a1b1862fcae800820ae717f8f078df90da4e106886fc9259e86ac23368ad7437" } }, "gpt-5-6-luna": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gpt-5-6-luna", "name": "GPT-5.6 Luna", "provider": "OpenAI", "releasedAt": "2026-07-09", "brand": { "src": "/model-providers/openai.svg", "monogram": "O", "raster": false }, "contextWindowTokens": 1050000, "weightsUrl": null }, "price": { "currency": "USD", "input": 0.2, "output": 1.2, "cached": 0.02, "inputCny": 1.3404600000000002, "outputCny": 8.04276, "cachedCny": 0.134046, "officialUrl": "https://developers.openai.com/api/docs/models", "note": null }, "overall": { "rank": 36, "score": 57.5, "pendingReason": null, "onBoard": false, "coverage": 1, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 27, "score": 59.2, "coverage": 1, "sourceCount": 9, "onBoard": true }, { "key": "reasoning", "name": "科研", "rank": 34, "score": 58.3, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 35, "score": 58.4, "coverage": 1, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 45, "score": 51.5, "coverage": 1, "sourceCount": 6, "onBoard": false } ], "metricCount": 35, "evidence": [ { "key": "coding", "name": "编程", "score": 59.2, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "22.6%", "reading": 67, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "67.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "82.9%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "44.5%", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "77.1%", "reading": 63.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.3" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "83.9%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "11.6%", "reading": 55.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 82, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.3" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "53.6%", "reading": 54, "bound": null, "weight": "full", "fallback": null, "sourceRank": 68, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "11.6%", "reading": 51.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 31, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "51.4" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 58.3, "items": [ { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "82.1%", "reading": 70, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 14, "sourceModelName": "gpt-5.6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-09T14:18:35.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "70.0" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "20.6%", "reading": 65.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.5" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "40.0%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "gpt-5.6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-09T13:36:49.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "85.6%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "39.5%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 77, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "0.0%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "14.4%", "reading": 52.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.6" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "21.0%", "reading": 52.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 53, "sourceModelName": "gpt-5.6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-28T20:59:07.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "52.2" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "3.3%", "reading": 50.9, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数 · 至多", "display": "50.9" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 58.4, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "78.0%", "reading": 65.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.1" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "55.0%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "50.2%", "reading": 62.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "62.2" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,342.7", "reading": 61.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 62, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.7" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,463.9", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "35.7%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "1.3%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "43.0%", "reading": 54.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "54.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "42.4%", "reading": 52.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 50, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.2" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 51.5, "items": [ { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "83.7%", "reading": 61.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 15, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "19.4%", "reading": 53, "bound": null, "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.0" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "41.0%", "reading": 49.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "gpt-5.6-luna_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T23:44:20.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "49.3" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "41.0%", "reading": 49, "bound": null, "weight": "full", "fallback": null, "sourceRank": 83, "sourceModelName": "openai/gpt-5.6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "49.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "72.6%", "reading": 48.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 60, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "48.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "60.1%", "reading": 47.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 60, "sourceModelName": "gpt-5.6-luna-max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "47.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "-10.3", "reading": 43, "bound": null, "weight": "full", "fallback": null, "sourceRank": 128, "sourceModelName": "gpt-5-6-luna", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "43.0" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 54.5, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,453.1", "reading": 51.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 75, "sourceModelName": "gpt-5.6-luna-xhigh", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "51.2" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gpt-5-6-luna", "api_url": "https://aihot.news/api/site/leaderboard/models/gpt-5-6-luna", "fetched_at": "2026-10-09T04:17:11.367Z", "raw_sha256": "ff436ee42578b15f4eb5048afc9d3315a2d1c9cfe9d4a12e06a9fa12c5136dee" } }, "grok-4-5": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "grok-4-5", "name": "Grok 4.5", "provider": "xAI", "releasedAt": "2026-07-08", "brand": { "src": "/model-providers/xai.svg", "monogram": "X", "raster": false }, "contextWindowTokens": 500000, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.3, "inputCny": 13.4046, "outputCny": 40.2138, "cachedCny": 2.01069, "officialUrl": "https://docs.x.ai/developers/pricing", "note": null }, "overall": { "rank": 31, "score": 58.3, "pendingReason": null, "onBoard": false, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 37, "score": 55.6, "coverage": 1, "sourceCount": 8, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 28, "score": 59.6, "coverage": 0.714, "sourceCount": 5, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 30, "score": 60.7, "coverage": 0.875, "sourceCount": 9, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 24, "score": 60.7, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 30, "evidence": [ { "key": "coding", "name": "编程", "score": 55.6, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "69.0%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "36.6%", "reading": 59.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.4" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "55.0%", "reading": 55.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.8" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "80.4%", "reading": 54.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "54.6" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "10.6%", "reading": 54.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 84, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.4" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "8.6%", "reading": 47.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "47.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "68.6%", "reading": 47.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 50, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "47.2" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 59.6, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "87.2%", "reading": 61.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "61.8" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "36.0%", "reading": 61.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "grok-4.5_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-07-08T23:54:22.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "15.4%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 79, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "42.7%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "57.2%", "reading": 55.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "grok-4.5_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-07-09T01:53:42.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "55.3" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 60.7, "items": [ { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "12.9%", "reading": 76, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 8, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "76.0" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "57.9%", "reading": 65.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.9" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "56.2%", "reading": 65.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "65.3" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,280.4", "reading": 59.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 70, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.2" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "73.0%", "reading": 58.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.3" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "35.0%", "reading": 58.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.1" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,389.2", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "48.3%", "reading": 55.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "55.5" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "43.3%", "reading": 53.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 45, "sourceModelName": "grok/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "53.4" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 60.7, "items": [ { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "25.3", "reading": 67.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.1" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "82.8%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "71.5%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "55.0%", "reading": 63.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 68, "sourceModelName": "x-ai/grok-4.5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "63.2" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "48.3%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "grok-4.5_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:29:30.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.3%", "reading": 54.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 82, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.3" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "15.0%", "reading": 46.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 48, "sourceModelName": "grok-4-5", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "46.5" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 57.8, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,465.8", "reading": 55.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 50, "sourceModelName": "grok-4.5", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "55.9" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/grok-4-5", "api_url": "https://aihot.news/api/site/leaderboard/models/grok-4-5", "fetched_at": "2026-10-09T04:17:10.447Z", "raw_sha256": "35daf005f130d16c7687bcb7ec3eb8810f4519109a888a38bd2d075aa649c6b3" } }, "gemini-3-1-pro-preview": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-3-1-pro-preview", "name": "Gemini 3.1 Pro Preview", "provider": "Google", "releasedAt": "2026-02-19", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "inputCny": 13.4046, "outputCny": 80.4276, "cachedCny": 1.3404600000000002, "officialUrl": "https://ai.google.dev/gemini-api/docs/pricing", "note": null }, "overall": { "rank": 40, "score": 56.1, "pendingReason": null, "onBoard": false, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 42, "score": 52, "coverage": 0.857, "sourceCount": 8, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 29, "score": 59.3, "coverage": 1, "sourceCount": 8, "onBoard": true }, { "key": "professional", "name": "专业办公", "rank": 42, "score": 54.9, "coverage": 0.875, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 16, "score": 63.3, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 33, "evidence": [ { "key": "coding", "name": "编程", "score": 52, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "58.7%", "reading": 60.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 17, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "76.5%", "reading": 55.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.5" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "4.0%", "reading": 47.6, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 100, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数 · 至多", "display": "47.6" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "17.3%", "reading": 46.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "46.2" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "32.0%", "reading": 45.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 67, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "45.9" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "6.7%", "reading": 42.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "42.5" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "2.5%", "reading": 40.9, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "40.9" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 59.3, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "49.0%", "reading": 70.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 12, "sourceModelName": "gemini-3.1-pro-preview_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-06T22:30:32.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "70.8" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "47.0%", "reading": 63.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.4" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "17.7%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "34.0%", "reading": 62.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "gemini-3.1-pro-preview_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-27T18:29:16.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "62.4" } ] }, { "sourceKey": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "display": "20.7%", "reading": 59, "bound": null, "weight": "full", "fallback": null, "sourceRank": 20, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.860Z", "measuredAt": null, "carriedForward": false, "sourceName": "MysteryMechanism", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.0" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "84.0%", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "59.6%", "reading": 56.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "gemini-3.1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-11T23:35:57.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "56.5" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "1.4%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 33, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 54.9, "items": [ { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "59.1%", "reading": 72.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 2, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "72.7" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "78.5%", "reading": 65.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 23, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "65.8" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "41.3%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "0.0%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 60, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "35.4%", "reading": 55.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 101, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.1" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "43.0%", "reading": 49.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "49.9" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "35.3%", "reading": 48.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "48.6" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "794.5", "reading": 34.1, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 156, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "34.1" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "456.1", "reading": 25.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 167, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "25.6" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 63.3, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "73.5%", "reading": 84.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 3, "sourceModelName": "gemini-3.1-pro-preview_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-10T21:35:39.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "84.9" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.1%", "reading": 76.9, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 5, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "76.9" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "31.9", "reading": 71.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.9" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "85.4%", "reading": 70.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 5, "sourceModelName": "gemini-3.1-pro-preview-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.8" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "82.0%", "reading": 58.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.8" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "15.6%", "reading": 47.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 47, "sourceModelName": "gemini-3-1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "47.4" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "31.0%", "reading": 38, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 131, "sourceModelName": "google/gemini-3.1-pro-preview", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "38.0" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 60, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,487", "reading": 63.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 18, "sourceModelName": "gemini-3.1-pro-preview", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "63.7" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-3-1-pro-preview", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-3-1-pro-preview", "fetched_at": "2026-10-09T04:17:10.372Z", "raw_sha256": "f5f74f30a90e2cea3b82edd798f0b5c0dd317f4f27ea508d81f09288179ca801" } }, "gemini-3-6-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-3-6-flash", "name": "Gemini 3.6 Flash", "provider": "Google", "releasedAt": "2026-07-21", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "inputCny": 5.026725, "outputCny": 25.133625000000002, "cachedCny": 0.5026725, "officialUrl": "https://ai.google.dev/gemini-api/docs/pricing", "note": null }, "overall": { "rank": 32, "score": 57.8, "pendingReason": null, "onBoard": false, "coverage": 0.929, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 35, "score": 56.8, "coverage": 1, "sourceCount": 7, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 38, "score": 57.3, "coverage": 0.857, "sourceCount": 7, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 39, "score": 57.5, "coverage": 0.875, "sourceCount": 8, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 23, "score": 61, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 30, "evidence": [ { "key": "coding", "name": "编程", "score": 56.8, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "64.0%", "reading": 57.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 44, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "57.7" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "77.9%", "reading": 57.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 54, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.2" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "30.9%", "reading": 56.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.1" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "53.4%", "reading": 53.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 69, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.7" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "7.1%", "reading": 50.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 89, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.7" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 57.3, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "40.0%", "reading": 64.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 24, "sourceModelName": "gemini-3.6-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-02T00:21:27.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.4" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "30.0%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "gemini-3.6-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-05T01:49:56.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "85.1%", "reading": 59.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 47, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "40.8%", "reading": 58.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.1" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "58.9%", "reading": 56.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 37, "sourceModelName": "gemini-3.6-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-02T00:36:45.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "56.2" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "10.6%", "reading": 55.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 100, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.3" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "4.3%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 23, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 57.5, "items": [ { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "53.2%", "reading": 65.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.4" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "56.3%", "reading": 63.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "63.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "53.0%", "reading": 63.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.5" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "46.9%", "reading": 58, "bound": null, "weight": "full", "fallback": null, "sourceRank": 36, "sourceModelName": "gemini-3.6-flash", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "58.0" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "3.3%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 39, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,285.9", "reading": 53.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 95, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.6" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "63.0%", "reading": 46.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 62, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "46.7" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "953", "reading": 45.8, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 118, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "45.8" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 61, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "66.2%", "reading": 76, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 11, "sourceModelName": "gemini-3.6-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:30:09.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "76.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.4%", "reading": 70.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.2" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "83.9%", "reading": 67.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "67.6" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "22.1", "reading": 64.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 52, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.8" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "80.0%", "reading": 55.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "14.4%", "reading": 45.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 54, "sourceModelName": "gemini-3-6-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "45.6" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "28.0%", "reading": 34.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 140, "sourceModelName": "google/gemini-3.6-flash", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "34.3" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 59.9, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,483", "reading": 62.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "gemini-3.6-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "62.2" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-3-6-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-3-6-flash", "fetched_at": "2026-10-09T04:17:10.333Z", "raw_sha256": "01221c413745bc2a726dd25f78ebf99bdcf571a1954937264144e8331dea5961" } }, "gemini-3-5-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "gemini-3-5-flash", "name": "Gemini 3.5 Flash", "provider": "Google", "releasedAt": "2026-05-19", "brand": { "src": "/model-providers/google.svg", "monogram": "G", "raster": false }, "contextWindowTokens": 1048576, "weightsUrl": null }, "price": { "currency": "USD", "input": 1.5, "output": 9, "cached": 0.15, "inputCny": 10.05345, "outputCny": 60.3207, "cachedCny": 1.005345, "officialUrl": "https://ai.google.dev/gemini-api/docs/pricing", "note": null }, "overall": { "rank": 39, "score": 56.7, "pendingReason": null, "onBoard": false, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 40, "score": 54.1, "coverage": 0.857, "sourceCount": 7, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 33, "score": 58.6, "coverage": 0.857, "sourceCount": 7, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 40, "score": 56.7, "coverage": 0.875, "sourceCount": 9, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 25, "score": 60.6, "coverage": 1, "sourceCount": 6, "onBoard": true } ], "metricCount": 31, "evidence": [ { "key": "coding", "name": "编程", "score": 54.1, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "78.2%", "reading": 57.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 45, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "57.5" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "53.9%", "reading": 54.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 65, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "26.7%", "reading": 53.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "53.5" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "48.7%", "reading": 52.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 55, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "52.1" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "6.6%", "reading": 50, "bound": null, "weight": "full", "fallback": null, "sourceRank": 92, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.0" } ] }, { "sourceKey": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "display": "6.1%", "reading": 43.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 37, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.846Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "43.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 58.6, "items": [ { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "50.0%", "reading": 71.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 11, "sourceModelName": "gemini-3.5-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-05-28T10:10:09.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "71.5" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "32.0%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "gemini-3.5-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-27T18:28:52.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "42.7%", "reading": 59.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "59.7" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "13.1%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 92, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "62.8%", "reading": 58.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "gemini-3.5-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-10T17:32:38.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "58.1" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "5.7%", "reading": 56.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "56.1" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "82.0%", "reading": 55.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 56.7, "items": [ { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "55.8%", "reading": 68.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 6, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "68.7" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "45.0%", "reading": 65.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "65.9" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "57.9%", "reading": 65.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 11, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.5" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "42.1%", "reading": 58.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 88, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.4" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "2.5%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,204.8", "reading": 50.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 106, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.4" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "64.9%", "reading": 48.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "48.7" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "872.8", "reading": 42.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 133, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "42.6" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "27.5%", "reading": 41.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 52, "sourceModelName": "gemini-3.5-flash", "configurationLabel": "High 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "41.4" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 60.6, "items": [ { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "66.2%", "reading": 76, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 11, "sourceModelName": "gemini-3.5-flash_high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:30:35.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "76.0" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "75.6%", "reading": 70.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 9, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "70.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "84.6%", "reading": 69, "bound": null, "weight": "full", "fallback": null, "sourceRank": 9, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "69.0" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "21.2", "reading": 64.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 54, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.2" } ] }, { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "18.3%", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "73.3%", "reading": 45.7, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 136, "sourceModelName": "gemini-3-5-flash", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "45.7" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "20.0%", "reading": 23.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 169, "sourceModelName": "google/gemini-3.5-flash", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "23.2" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 58.6, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,477.4", "reading": 60.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "gemini-3.5-flash-high", "configurationLabel": "High 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "60.2" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/gemini-3-5-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/gemini-3-5-flash", "fetched_at": "2026-10-09T04:17:12.490Z", "raw_sha256": "ca0ad433d75d9575e2a4ce0ee963ef47276a231f4bcbd90bfc80b3d18a2663cc" } }, "qwen-3-7-max": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "qwen-3-7-max", "name": "Qwen3.7 Max", "provider": "Alibaba", "releasedAt": "2026-05-19", "brand": { "src": "/model-providers/qianwen.svg", "monogram": "A", "raster": false }, "contextWindowTokens": 1000000, "weightsUrl": null }, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 2.4, "inputCny": 12, "outputCny": 36, "cachedCny": 2.4, "officialUrl": "https://help.aliyun.com/zh/model-studio/model-pricing", "note": null }, "overall": { "rank": 43, "score": 52.8, "pendingReason": null, "onBoard": false, "coverage": 0.75, "missingDimensions": [ "真人盲选" ] }, "categories": [ { "key": "coding", "name": "编程", "rank": 47, "score": 49.5, "coverage": 0.714, "sourceCount": 5, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 39, "score": 56.4, "coverage": 0.714, "sourceCount": 6, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 47, "score": 50.2, "coverage": 0.75, "sourceCount": 8, "onBoard": false }, { "key": "knowledge", "name": "知识问答", "rank": 29, "score": 59.7, "coverage": 1, "sourceCount": 5, "onBoard": true } ], "metricCount": 25, "evidence": [ { "key": "coding", "name": "编程", "score": 49.5, "items": [ { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "74.2%", "reading": 53, "bound": null, "weight": "full", "fallback": null, "sourceRank": 60, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "53.0" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "47.7%", "reading": 51.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 58, "sourceModelName": "alibaba/qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "51.7" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "49.5%", "reading": 48.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 108, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "48.9" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "1.5%", "reading": 47.6, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 115, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数 · 至多", "display": "47.6" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "13.1%", "reading": 41.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 57, "sourceModelName": "alibaba/qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "41.9" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 56.4, "items": [ { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "32.0%", "reading": 61.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-07-28T19:23:45.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "61.1" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "64.6%", "reading": 59, "bound": null, "weight": "full", "fallback": null, "sourceRank": 32, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-06-13T03:08:22.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "59.0" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "13.4%", "reading": 58.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 90, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.8" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "40.5%", "reading": 57.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 74, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.8" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "83.3%", "reading": 56.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 52, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.9" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "19.0%", "reading": 44.5, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 99, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-07T00:46:56.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "44.5" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 50.2, "items": [ { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "71.8%", "reading": 56.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "56.8" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "1.7%", "reading": 56.3, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 46, "sourceModelName": "alibaba/qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "56.3" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "47.8%", "reading": 54.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 48, "sourceModelName": "alibaba/qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "54.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "23.0%", "reading": 47.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 123, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "47.9" } ] }, { "sourceKey": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/medcode", "display": "38.8%", "reading": 47.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 69, "sourceModelName": "alibaba/qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.939Z", "measuredAt": null, "carriedForward": false, "sourceName": "MedCode", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "47.6" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,131.2", "reading": 47.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 116, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "47.4" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "915.2", "reading": 44.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 125, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "44.3" } ] }, { "sourceKey": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "display": "18.8%", "reading": 42.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.154Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Analyst Agent", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "42.3" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 59.7, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.0%", "reading": 68, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "68.0" } ] }, { "sourceKey": "bullshitbench-v2", "name": "BullshitBench v2", "operator": "BullshitBench", "brand": { "src": null, "monogram": "P", "raster": false }, "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "display": "57.0%", "reading": 65.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 66, "sourceModelName": "qwen/qwen3.7-max", "configurationLabel": "xHigh 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T20:50:19.000Z", "verifiedAt": "2026-10-09T00:05:44.778Z", "measuredAt": null, "carriedForward": false, "sourceName": "BullshitBench v2", "usage": "BullshitBench", "components": [ { "label": "AIHOT 读数", "display": "65.3" } ] }, { "sourceKey": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "display": "55.8%", "reading": 64.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.685Z", "measuredAt": "2026-08-27T19:34:09.000Z", "carriedForward": false, "sourceName": "SimpleQA Verified", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "64.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "79.7%", "reading": 59.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 21, "sourceModelName": "qwen3.7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "13.5", "reading": 58.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 71, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "58.9" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.0%", "reading": 53.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 90, "sourceModelName": "qwen3-7-max", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "53.8" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-program", "name": "ProgramBench" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA" }, { "key": "arena-text", "name": "Arena Text" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/qwen-3-7-max", "api_url": "https://aihot.news/api/site/leaderboard/models/qwen-3-7-max", "fetched_at": "2026-10-09T04:17:11.898Z", "raw_sha256": "f34e13d8c68980b3830ac8ab32f47d75e7faae2edf3bc0d57e5706b760187b1e" } }, "glm-5-3-flash": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "glm-5-3-flash", "name": "GLM-5.3 Flash", "provider": "Z.ai", "releasedAt": "2026-08-26", "brand": { "src": "/model-providers/z-ai.svg", "monogram": "Z", "raster": false }, "contextWindowTokens": 1310720, "weightsUrl": "https://huggingface.co/zai-org/GLM-5.3-Flash" }, "price": { "currency": "CNY", "input": 0.8, "output": 2.8, "cached": 0.23, "inputCny": 0.8, "outputCny": 2.8, "cachedCny": 0.23, "officialUrl": "https://www.bigmodel.cn/pricing", "note": null }, "overall": { "rank": 38, "score": 57.1, "pendingReason": null, "onBoard": false, "coverage": 0.893, "missingDimensions": [] }, "categories": [ { "key": "coding", "name": "编程", "rank": 34, "score": 56.9, "coverage": 1, "sourceCount": 8, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 42, "score": 53.7, "coverage": 0.857, "sourceCount": 8, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 20, "score": 62.7, "coverage": 0.875, "sourceCount": 8, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 37, "score": 54.9, "coverage": 0.8, "sourceCount": 4, "onBoard": false } ], "metricCount": 30, "evidence": [ { "key": "coding", "name": "编程", "score": 56.9, "items": [ { "sourceKey": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/programbench", "display": "0.0%", "reading": 84, "bound": "max", "weight": "weak", "fallback": null, "sourceRank": 31, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.822Z", "measuredAt": null, "carriedForward": false, "sourceName": "ProgramBench", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "84.0" } ] }, { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "32.8%", "reading": 66.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 38, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.2" } ] }, { "sourceKey": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/cyber", "display": "85.7%", "reading": 62.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.898Z", "measuredAt": null, "carriedForward": false, "sourceName": "CyberBench Patch v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.5" } ] }, { "sourceKey": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "display": "16.0%", "reading": 59.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 14, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.882Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench 1–100", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "59.6" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "79.0%", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 22, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "51.6%", "reading": 51.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 85, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "51.5" } ] }, { "sourceKey": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "display": "20.5%", "reading": 48.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 49, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.776Z", "measuredAt": null, "carriedForward": false, "sourceName": "Code Migration", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "48.9" } ] }, { "sourceKey": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "display": "30.8%", "reading": 45.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 69, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.877Z", "measuredAt": null, "carriedForward": false, "sourceName": "Vibe Code Bench v1.1", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "45.3" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 53.7, "items": [ { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "15.4%", "reading": 60.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 79, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "60.9" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "39.9%", "reading": 57.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 75, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "57.2" } ] }, { "sourceKey": "epoch-frontiermath", "name": "FrontierMath T1–3", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "display": "55.8%", "reading": 54.7, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "glm-5.3-flash_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.635Z", "measuredAt": "2026-08-27T13:29:52.000Z", "carriedForward": false, "sourceName": "FrontierMath T1–3", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "54.7" } ] }, { "sourceKey": "vals-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "display": "2.9%", "reading": 54.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 30, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.919Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "54.7" } ] }, { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "77.6%", "reading": 51, "bound": null, "weight": "full", "fallback": null, "sourceRank": 61, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.0" } ] }, { "sourceKey": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "display": "4.8%", "reading": 50.9, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 35, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.199Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench Science", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数 · 至多", "display": "50.9" } ] }, { "sourceKey": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "display": "8.0%", "reading": 40.7, "bound": "max", "weight": "full", "fallback": null, "sourceRank": 113, "sourceModelName": "glm-5.3-flash_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.677Z", "measuredAt": "2026-08-28T14:38:27.000Z", "carriedForward": false, "sourceName": "Mystery Game Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数 · 至多", "display": "40.7" } ] }, { "sourceKey": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "display": "14.0%", "reading": 36.3, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 113, "sourceModelName": "glm-5.3-flash_max", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-09-29T22:00:34.000Z", "verifiedAt": "2026-10-09T00:05:26.667Z", "measuredAt": "2026-08-26T22:26:55.000Z", "carriedForward": false, "sourceName": "Chess Puzzles", "usage": "Epoch AI", "components": [ { "label": "AIHOT 读数", "display": "36.3" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.7, "items": [ { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,646.9", "reading": 67.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.9" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "60.4%", "reading": 67.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.1" } ] }, { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,453.7", "reading": 66.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 43, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "66.2" } ] }, { "sourceKey": "vals-finance-agent", "name": "Finance Agent", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "display": "57.9%", "reading": 65.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 12, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.756Z", "measuredAt": null, "carriedForward": false, "sourceName": "Finance Agent", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "65.5" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "76.4%", "reading": 62.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 34, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.8" } ] }, { "sourceKey": "mercor-apex-agents", "name": "APEX-Agents", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "display": "52.8%", "reading": 62.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "glm-5-3-flash", "configurationLabel": "Max 推理 · loop_truncated_tools_agent · apex-agents:v1.1:loop_truncated_tools_agent", "selectionReason": "该来源对所有模型使用同一受控系统；按预先固定的推理档位优先级选择，配置明细保留，成绩归到基础模型。", "upstreamAt": "2026-09-08T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:29.879Z", "measuredAt": null, "carriedForward": false, "sourceName": "APEX-Agents", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "62.6" } ] }, { "sourceKey": "vals-harvey-lab", "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "officialUrl": "https://www.vals.ai/benchmarks/hlab", "display": "6.7%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 26, "sourceModelName": "zai/glm-5.3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": "2026-10-07T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:42.854Z", "measuredAt": null, "carriedForward": false, "sourceName": "Harvey 法律 Agent 基准", "usage": "Vals AI", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "mercor-tau3-banking", "name": "τ³-Banking", "operator": "Mercor", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "display": "37.8%", "reading": 61.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "glm-5-3-flash", "configurationLabel": "Max 推理", "selectionReason": "按预先固定的规则，采用该来源可用的最高第一方推理档位；选择不参考跑分高低。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:29.891Z", "measuredAt": null, "carriedForward": false, "sourceName": "τ³-Banking", "usage": "Mercor", "components": [ { "label": "AIHOT 读数", "display": "61.5" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 54.9, "items": [ { "sourceKey": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "display": "51.1%", "reading": 83.6, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 12, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.165Z", "measuredAt": null, "carriedForward": false, "sourceName": "MLCR-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "83.6" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "77.3%", "reading": 55.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 62, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "55.8" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "80.0%", "reading": 55.4, "bound": null, "weight": "full", "fallback": null, "sourceRank": 64, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.4" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "7.5", "reading": 54.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 76, "sourceModelName": "glm-5-3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "52.8%", "reading": 38.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 62, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "38.4" } ] } ] }, { "key": "preference", "name": "真人盲选", "score": 57.6, "items": [ { "sourceKey": "arena-text", "name": "Arena Text", "operator": "LMArena", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "display": "1,473", "reading": 58.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 41, "sourceModelName": "glm-5.3-flash", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-02T00:00:00.000Z", "verifiedAt": "2026-10-09T00:05:21.217Z", "measuredAt": null, "carriedForward": false, "sourceName": "Arena Text", "usage": "LMArena", "components": [ { "label": "AIHOT 读数", "display": "58.5" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-medcode", "name": "MedCode" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/glm-5-3-flash", "api_url": "https://aihot.news/api/site/leaderboard/models/glm-5-3-flash", "fetched_at": "2026-10-09T04:17:11.886Z", "raw_sha256": "1f283dd1b894d3f9aa794c08aca20ad9eb7df1097c7d03f5cdf00570656fd61f" } }, "qwen-3-8-flash-next": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "historical": false, "model": { "slug": "qwen-3-8-flash-next", "name": "Qwen3.8 Flash Next", "provider": "Alibaba", "releasedAt": "2026-08-26", "brand": { "src": "/model-providers/qianwen.svg", "monogram": "A", "raster": false }, "contextWindowTokens": null, "weightsUrl": null }, "price": { "currency": "CNY", "input": 0.8, "output": 2.7, "cached": 0.1, "inputCny": 0.8, "outputCny": 2.7, "cachedCny": 0.1, "officialUrl": "https://help.aliyun.com/zh/model-studio/qwen3-8-flash", "note": null }, "overall": { "rank": 34, "score": 57.6, "pendingReason": null, "onBoard": false, "coverage": 0.464, "missingDimensions": [ "真人盲选" ] }, "categories": [ { "key": "coding", "name": "编程", "rank": 36, "score": 56, "coverage": 0.429, "sourceCount": 3, "onBoard": false }, { "key": "reasoning", "name": "科研", "rank": 35, "score": 57.9, "coverage": 0.429, "sourceCount": 3, "onBoard": false }, { "key": "professional", "name": "专业办公", "rank": 24, "score": 62.2, "coverage": 0.375, "sourceCount": 4, "onBoard": true }, { "key": "knowledge", "name": "知识问答", "rank": 35, "score": 55.4, "coverage": 0.8, "sourceCount": 3, "onBoard": false } ], "metricCount": 14, "evidence": [ { "key": "coding", "name": "编程", "score": 56, "items": [ { "sourceKey": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "display": "25.3%", "reading": 63.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 50, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.889Z", "measuredAt": null, "carriedForward": false, "sourceName": "Terminal-Bench 4.0", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "63.1" } ] }, { "sourceKey": "livebench-coding", "name": "LiveBench Coding", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "display": "72.6%", "reading": 51.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 29, "sourceModelName": "qwen3.8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.528Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Coding", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.2" } ] }, { "sourceKey": "artificial-analysis-scicode", "name": "SciCode", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "display": "50.6%", "reading": 50.2, "bound": null, "weight": "full", "fallback": null, "sourceRank": 95, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.901Z", "measuredAt": null, "carriedForward": false, "sourceName": "SciCode", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "50.2" } ] } ] }, { "key": "reasoning", "name": "科研", "score": 57.9, "items": [ { "sourceKey": "livebench-reasoning", "name": "LiveBench Reasoning", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "display": "87.4%", "reading": 62.1, "bound": null, "weight": "full", "fallback": null, "sourceRank": 42, "sourceModelName": "qwen3.8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.537Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Reasoning", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "62.1" } ] }, { "sourceKey": "artificial-analysis-critpt", "name": "CritPt", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "display": "11.1%", "reading": 56, "bound": null, "weight": "full", "fallback": null, "sourceRank": 98, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.928Z", "measuredAt": null, "carriedForward": false, "sourceName": "CritPt", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "56.0" } ] }, { "sourceKey": "artificial-analysis-hle", "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "display": "38.0%", "reading": 55.6, "bound": null, "weight": "full", "fallback": null, "sourceRank": 87, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.914Z", "measuredAt": null, "carriedForward": false, "sourceName": "Humanity's Last Exam", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "55.6" } ] } ] }, { "key": "professional", "name": "专业办公", "score": 62.2, "items": [ { "sourceKey": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "display": "1,582.6", "reading": 71.5, "bound": null, "weight": "full", "fallback": null, "sourceRank": 16, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.016Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Briefcase", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "71.5" } ] }, { "sourceKey": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "display": "1,632.4", "reading": 67.3, "bound": null, "weight": "full", "fallback": null, "sourceRank": 19, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:20.941Z", "measuredAt": null, "carriedForward": false, "sourceName": "GDPval-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "67.3" } ] }, { "sourceKey": "artificial-analysis-automationbench", "name": "AutomationBench-AA", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "display": "55.9%", "reading": 64.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 51, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.084Z", "measuredAt": null, "carriedForward": false, "sourceName": "AutomationBench-AA", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "64.9" } ] }, { "sourceKey": "livebench-data-analysis", "name": "LiveBench Data Analysis", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "display": "74.2%", "reading": 59.9, "bound": null, "weight": "full", "fallback": null, "sourceRank": 40, "sourceModelName": "qwen3.8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.544Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Data Analysis", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "59.9" } ] } ] }, { "key": "knowledge", "name": "知识问答", "score": 55.4, "items": [ { "sourceKey": "livebench-writing", "name": "LiveBench IF", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "77.1%", "reading": 73.2, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 28, "sourceModelName": "qwen3.8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench IF", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "73.2" } ] }, { "sourceKey": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "display": "79.7%", "reading": 54.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 73, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.132Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-LCR", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "54.8" } ] }, { "sourceKey": "livebench-writing", "name": "LiveBench Language", "operator": "LiveBench", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "display": "74.6%", "reading": 51.8, "bound": null, "weight": "full", "fallback": null, "sourceRank": 28, "sourceModelName": "qwen3.8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": "2026-10-07T20:15:35.000Z", "verifiedAt": "2026-10-09T00:05:24.517Z", "measuredAt": null, "carriedForward": false, "sourceName": "LiveBench Language", "usage": "LiveBench", "components": [ { "label": "AIHOT 读数", "display": "51.8" } ] }, { "sourceKey": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "display": "-9.7", "reading": 43.4, "bound": null, "weight": "capped", "fallback": null, "sourceRank": 124, "sourceModelName": "qwen3-8-flash-next", "configurationLabel": "来源默认配置", "selectionReason": "该来源没有区分推理档位，采用其官方默认配置。", "upstreamAt": null, "verifiedAt": "2026-10-09T00:05:21.106Z", "measuredAt": null, "carriedForward": false, "sourceName": "AA-Omniscience", "usage": "Artificial Analysis", "components": [ { "label": "AIHOT 读数", "display": "43.4" } ] } ] } ], "excluded": [], "unmeasured": [ { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals" }, { "key": "vals-migration", "name": "Code Migration" }, { "key": "vals-program", "name": "ProgramBench" }, { "key": "vals-vibe-code", "name": "Vibe Code Bench v1.1" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3" }, { "key": "epoch-chess", "name": "Chess Puzzles" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA" }, { "key": "vals-finance-agent", "name": "Vals Finance Agent" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1" }, { "key": "vals-harvey-lab", "name": "Harvey Legal Agent Benchmark" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor" }, { "key": "vals-medcode", "name": "MedCode" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提" }, { "key": "arena-text", "name": "Arena Text" } ], "comparisons": [], "provenance": { "source_url": "https://aihot.news/leaderboard/qwen-3-8-flash-next", "api_url": "https://aihot.news/api/site/leaderboard/models/qwen-3-8-flash-next", "fetched_at": "2026-10-09T04:17:12.573Z", "raw_sha256": "310b5a969eff1295e417cba8f95ffe51c62e67d091fbea2bf05784eae85752cb" } } }, "source_details": { "artificial-analysis-terminal-bench": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0 · AA", "operator": "Artificial Analysis", "description": "在命令行环境里独立完成一整件工程任务。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Terminal-Bench 4.0（Artificial Analysis 复跑）", "area": "编程", "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "what": "取 Artificial Analysis 智能指数中 Terminal-Bench 4.0 的通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域，与 Vals 复跑的 Terminal-Bench 4.0 同属一类，两家合起来算一票。", "limits": "终端任务对运行环境和时限设置敏感；厂商接口回退的运行照常计入并标注（AA 不公布受影响比例）。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.889Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-terminal-bench", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-terminal-bench", "fetched_at": "2026-10-09T04:16:22.541Z", "raw_sha256": "6d1cdc7833c9920be8f58be74953e0ca6ccacf5dd64a5932515dd402fd045fdd" } }, "vals-terminal-bench-4": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals", "operator": "Vals AI", "description": "Vals 用同一套 Agent 流程给所有模型复跑 Terminal-Bench 4.0。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Terminal-Bench 4.0（Vals AI 复跑）", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域，与 AA 复跑的 Terminal-Bench 4.0 同属一类，两家合起来算一票。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.846Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-terminal-bench-4", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-terminal-bench-4", "fetched_at": "2026-10-09T04:16:22.525Z", "raw_sha256": "95515dfd9d95e0e61a8ee1f269f541e9bb54f7c4d70f507f695a1cebf4c35eac" } }, "livebench-coding": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "livebench-coding", "name": "LiveBench · 编程综合", "operator": "LiveBench", "description": "Coding + Agentic Coding 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 编程综合", "area": "编程", "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "what": "读取 Coding 和 Agentic Coding 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：只计 Coding 一项（编程领域的写代码一类）。Agentic Coding 对部分厂商使用原生工具调用提示，流程不统一，暂不计分。", "limits": "两项均分衡量组合表现，不能当成 Coding 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-07T20:15:35.000Z", "syncedAt": "2026-10-09T00:05:24.528Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/livebench-coding", "api_url": "https://aihot.news/api/site/leaderboard/sources/livebench-coding", "fetched_at": "2026-10-09T04:16:22.626Z", "raw_sha256": "d7317af231758a917f08a8a6d6a576b1c7010cfba457a9a0d8227629adf9ef73" } }, "artificial-analysis-scicode": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-scicode", "name": "SciCode · AA", "operator": "Artificial Analysis", "description": "把物理、化学、生物等科研问题写成能运行的计算代码。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "SciCode（Artificial Analysis 复跑）", "area": "编程", "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "what": "取 Artificial Analysis 智能指数中 SciCode 的通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的科学计算一类。", "limits": "题目来自公开数据集，可能存在训练污染；题集正在接受公开复核。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.901Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-scicode", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-scicode", "fetched_at": "2026-10-09T04:16:23.657Z", "raw_sha256": "5117f392bae92f1ad69ea7a926410da267ae65ddbab106996de2578c7dbd388b" } }, "vals-migration": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "description": "把整套程序从一种语言或框架迁移到另一种。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Code Migration", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的程序迁移一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.776Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-migration", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-migration", "fetched_at": "2026-10-09T04:16:24.122Z", "raw_sha256": "38b1c3f20fb3e74558cf802afe31e0c3a207c84f5508e77e7c9b6a2fdf6faf58" } }, "vals-program": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "description": "只给可执行文件，从零重写出行为相同的程序。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "ProgramBench", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/programbench", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的程序重建一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.822Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-program", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-program", "fetched_at": "2026-10-09T04:16:24.872Z", "raw_sha256": "550995af05058e0f287b250f480727389f18ff738094fb35e6b95b4ec240914d" } }, "vals-vibe-code": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "description": "从需求搭建网页应用，检查实际功能。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vibe Code Bench v1.1", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "what": "50 个应用开发场景，统一 OpenHands，浏览器评测按验收步骤检查功能。", "usage": "计分：应用开发；与 Vibe Code Bench 1–100 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 OpenHands 成绩；厂商专用 Agent 不计。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.877Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-vibe-code", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-vibe-code", "fetched_at": "2026-10-09T04:16:24.848Z", "raw_sha256": "e4ee92d2c2d1a4a9ff74b2a1f40a829436ea3f56d3ec21cf4784c83ae20a097a" } }, "vals-vcb-1-100": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "description": "在已有应用上连续开发，同时保持旧功能。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vibe Code Bench 1–100", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "what": "50 个连续开发测试场景；统一 OpenHands、每场景 10 小时预算，浏览器验证新增功能与回归。", "usage": "计分：应用开发；与 Vibe Code Bench 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "版本 1.0；连续开发总体成绩。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.882Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-vcb-1-100", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-vcb-1-100", "fetched_at": "2026-10-09T04:16:24.798Z", "raw_sha256": "31c0954bd1dd018d0aadcebb2dfe690ea11cb8750888d7cd1db6ace4fd0bd00c" } }, "vals-cyber-patch": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "description": "修补真实开源项目的安全漏洞。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "CyberBench Patch v1.1", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/cyber", "what": "56 个 Patch 任务，统一 mini-swe-agent；编译、原漏洞复测及隐藏输入与维护者修复对照。", "usage": "计分：漏洞修补一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 Patch，不混入 PoC 或总分；回退比例按 Patch 的 56 题计算。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.898Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-cyber-patch", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-cyber-patch", "fetched_at": "2026-10-09T04:16:26.013Z", "raw_sha256": "9149183c9d96d1e15fd0f9e2f377b41a8ae9065b1b73fa724286ed9fade26be1" } }, "deepswe-v1-1": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "deepswe-v1-1", "name": "DeepSWE v1.1", "operator": "DataCurve", "description": "在统一的编程助手环境里改仓库、修缺陷，比较的是写代码的模型本身。", "status": "observing", "brand": { "src": "/leaderboard-sources/deepswe.svg", "monogram": "D", "raster": false }, "fullName": "DeepSWE v1.1", "area": "编程", "officialUrl": "https://deepswe.datacurve.ai/", "what": "只取 pass@1；统一 harness 内按预先固定的推理档位优先级选代表配置。", "usage": "观察中：9 月 3 日后没有再加入新模型，且 Epoch 指出部分题目判分有缺陷；恢复更新并修好后再计分。", "limits": "仍是模型＋mini-swe-agent 的系统表现；大版本变化必须隔离快照。", "license": "官方公开结构化结果；仅管理员私有观察，公开前复核许可", "attribution": "成绩由 DataCurve 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/deepswe-v1-1", "api_url": "https://aihot.news/api/site/leaderboard/sources/deepswe-v1-1", "fetched_at": "2026-10-09T04:16:26.002Z", "raw_sha256": "452642759b7e0afbe094c262062e1ea51943ddc30c319f3310b49fdd38173097" } }, "epoch-mirrorcode": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-mirrorcode", "name": "MirrorCode", "operator": "Epoch AI", "description": "长时间的软件复现", "status": "observing", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "MirrorCode", "area": "编程", "officialUrl": "https://epoch.ai/benchmarks/mirrorcode", "what": "已发布数据仅少量型号，且运行时间很长；样本和运行条件不足以形成分类共识。", "usage": "观察中：只测了少数模型，测满 20 个模型再计分。", "limits": "只测了少数模型，刻度定不准。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-mirrorcode", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-mirrorcode", "fetched_at": "2026-10-09T04:16:25.999Z", "raw_sha256": "2748c5df342f7e1cea9b27b9e87322c0e7bec609f140233746ac703bb2596724" } }, "taptap-maker": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "taptap-maker", "name": "TapTap Maker", "operator": "TapTap Maker", "description": "在真实游戏引擎里用 Lua 写出能跑起来的功能。分数由引擎执行判定，不用另一个模型打分。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/taptap-maker.svg", "monogram": "T", "raster": false }, "fullName": "TapTap Maker Benchmark", "area": "编程", "officialUrl": "https://maker.taptap.cn/leaderboard/", "what": "只取官方 main_board 的 L2 通过率；成本、速度、Held-out 和分类分不再次投票。同一基础模型按固定推理档位优先级选代表行，选择不看分数。", "usage": "仅供参考：现有结果混合不同 agent driver，不能作为相同执行流程下的底层型号比较。", "limits": "单引擎、单语言，不能推广到其他技术栈；不同模型可能使用不同 agent 驱动；闭源 harness 不能逐题复算。 当前采用的执行驱动不同，不计入单型号排序。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以易玩／TapTap 条款为准", "attribution": "成绩由 TapTap Maker 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-09-10T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:29.871Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/taptap-maker", "api_url": "https://aihot.news/api/site/leaderboard/sources/taptap-maker", "fetched_at": "2026-10-09T04:16:25.948Z", "raw_sha256": "391214261282b10d41b2771d1a8d3d4e32d893f09de0d493c2e8ab96a1900b1a" } }, "terminal-bench-4": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "terminal-bench-4", "name": "Terminal-Bench 4", "operator": "Harbor / Terminal-Bench", "description": "保留模型搭配不同 Agent 在终端任务中的原始成绩，供交叉核对。", "status": "reference_only", "brand": { "src": null, "monogram": "H", "raster": false }, "fullName": "Terminal-Bench 4 · 系统参考", "area": "编程", "officialUrl": "https://www.tbench.ai/", "what": "固定 Terminal-Bench 4.0，分别保留模型、Agent 版本、推理档位和运行次数；不把不同工具系统解释为统一模型条件。", "usage": "原始成绩参考：锁定官方提交清单与数据版本后自动采集，逐条保留不同系统配置，不选模型代表分。只供交叉核对，不进入综合或编程排名。", "limits": "不同模型使用 Claude Code、Codex、GrokBuild 或 mini-SWE 等不同工具，只供参考；AA 和 Vals 各自统一复跑的 Terminal-Bench 4.0 已计分。", "license": "Apache 2.0 · 官方 harbor-framework/terminal-bench 仓库内的成绩提交；保留署名、协议与修改说明。", "attribution": "成绩由 Harbor / Terminal-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": "2026-10-06T23:19:56.000Z", "syncedAt": "2026-10-09T00:05:42.980Z", "collected": true, "systemRows": true, "rowsNote": "下列为模型搭配不同 Agent 的系统成绩，运行条件不同，仅供参考。每项最多展示 30 条配置，匿名测试型号不展示。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/terminal-bench-4", "api_url": "https://aihot.news/api/site/leaderboard/sources/terminal-bench-4", "fetched_at": "2026-10-09T04:16:27.064Z", "raw_sha256": "59d354b98e3ce683d780baca17cda2f894f469e35f0894df2faf1924f35df0c2" } }, "hyper-tau": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "hyper-tau", "name": "Hyper-τ-bench", "operator": "Sierra", "description": "根据业务资料构建并测试能工作的客户服务代理。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "Hyper-τ-bench", "area": "编程", "officialUrl": "https://sierra.ai/blog/hyper-t-bench-evaluating-agents-that-build-agents", "what": "衡量代理开发与调试，和直接执行客户服务的 τ³-Banking 分开；固定开发工具、运行时长与验收环境，不能跨系统挑最高成绩。", "usage": "观察中：可补编程与业务系统开发，先完成原始成绩、许可和固定运行条件审查；不因新发布就自动计分。", "limits": "观察中：可补编程与业务系统开发，先完成原始成绩、许可和固定运行条件审查；不因新发布就自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Sierra 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/hyper-tau", "api_url": "https://aihot.news/api/site/leaderboard/sources/hyper-tau", "fetched_at": "2026-10-09T04:16:27.147Z", "raw_sha256": "8798d8d5af24f72fabce16bb33e3a2a323c06044a793b9dfdb8ce554c3bcb066" } }, "swe-rebench": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "swe-rebench", "name": "SWE-rebench", "operator": "Nebius", "description": "持续收集真实仓库问题，比较模型在多种编程语言中的软件修复能力。", "status": "observing", "brand": { "src": null, "monogram": "N", "raster": false }, "fullName": "SWE-rebench", "area": "编程", "officialUrl": "https://swe-rebench.com/", "what": "固定任务窗口、ReAct 环境和 text/tools 交互协议，采用多次运行的平均解决率与标准误；不把完整 Agent 产品与基础模型混为同一对象。", "usage": "候选观察：真实仓库任务值得补充，等待可验证运行日期、稳定成绩协议与再展示边界；尚未启用自动采集或计分。", "limits": "滚动题目、运行环境和任务窗口影响可比性；网站近期维护不等于所有模型重新测试。", "license": "官方 Hugging Face CC BY 4.0 数据集是任务题库；尚未确认网页最终成绩的稳定导出与再展示边界。", "attribution": "成绩由 Nebius 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/swe-rebench", "api_url": "https://aihot.news/api/site/leaderboard/sources/swe-rebench", "fetched_at": "2026-10-09T04:16:27.827Z", "raw_sha256": "9518e9886592df5493d14619d7fc953c82f41f5a62c1af17e3e5d666b182bbc6" } }, "lhtb": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "lhtb", "name": "LHTB", "operator": "Tencent HY / Lehigh 等", "description": "在较长时间内持续使用终端，完成复杂工程与工具任务。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "Long-Horizon Terminal-Bench", "area": "编程", "officialUrl": "https://zli12321.github.io/LHTB/leaderboard.html", "what": "90 分钟、2 小时和 3 小时预算分别比较；修复前后判题协议不能混排，也不能把不同 Agent 的最好结果当统一模型成绩。", "usage": "候选观察：等待明确标注修复协议的新批次，现有旧分不自动采集到公开榜，也不计分；不据此断言每条旧结果都利用了漏洞。", "limits": "官方披露过判题信息泄漏；当前不能证明已有成绩属于隔离判题后的同一协议。部分任务与其他专业评测的来源重叠待核。", "license": "Apache 2.0 · 官方 IntelligenceLab/LHTB-leaderboard 成绩数据集；保留署名和变更说明。", "attribution": "成绩由 Tencent HY / Lehigh 等 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/lhtb", "api_url": "https://aihot.news/api/site/leaderboard/sources/lhtb", "fetched_at": "2026-10-09T04:16:27.842Z", "raw_sha256": "8110b045fee89580197e0aef1974051fa492ba643d6273b13916d6767be8dcfa" } }, "frontiercode": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "frontiercode", "name": "FrontierCode", "operator": "Cognition", "description": "真实仓库任务的质量、测试与修改范围", "status": "observing", "brand": { "src": null, "monogram": "C", "raster": false }, "fullName": "FrontierCode", "area": "编程", "officialUrl": "https://cognition.com/frontiercode", "what": "固定 v1.1 和官方运行条件；质量评分和测试成功共同构成一项任务证据。", "usage": "观察中：尚无确认稳定的成绩导出与数据使用边界。", "limits": "观察中：尚无确认稳定的成绩导出与数据使用边界。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Cognition 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/frontiercode", "api_url": "https://aihot.news/api/site/leaderboard/sources/frontiercode", "fetched_at": "2026-10-09T04:16:28.291Z", "raw_sha256": "17695d55bad5c6da54dbab14705cab33657434d0a9a0a368aefad3a9cd054a19" } }, "frontierswe-v2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "frontierswe-v2", "name": "FrontierSWE v2", "operator": "Proximal Labs", "description": "长时间运行的工程实现与研究任务", "status": "observing", "brand": { "src": null, "monogram": "P", "raster": false }, "fullName": "FrontierSWE v2", "area": "编程", "officialUrl": "https://www.frontierswe.com/", "what": "只比较 v2 的统一 Proximus harness，不与旧版和 best-of 不同条件混合。", "usage": "观察中：长任务新型号覆盖仍较少，稳定原始数据入口和运行配置需要补齐。", "limits": "观察中：长任务新型号覆盖仍较少，稳定原始数据入口和运行配置需要补齐。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Proximal Labs 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/frontierswe-v2", "api_url": "https://aihot.news/api/site/leaderboard/sources/frontierswe-v2", "fetched_at": "2026-10-09T04:16:29.049Z", "raw_sha256": "011f7bbb1cbc9b9839ce3851e41cd5dfde127cf98c02ba23ccc24aa79d4f0332" } }, "weirdml": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "weirdml", "name": "WeirdML", "operator": "WeirdML", "description": "在陌生数据上实现机器学习方案", "status": "observing", "brand": { "src": null, "monogram": "W", "raster": false }, "fullName": "WeirdML", "area": "编程", "officialUrl": "https://htihle.github.io/weirdml.html", "what": "按实际任务清单和固定 GPU 时间比较，保留五轮迭代的计分协议。", "usage": "观察中：官方 CSV 可读，但成绩许可和逐型号运行时间尚未完整核实。", "limits": "观察中：官方 CSV 可读，但成绩许可和逐型号运行时间尚未完整核实。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 WeirdML 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "coding", "name": "编程" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/weirdml", "api_url": "https://aihot.news/api/site/leaderboard/sources/weirdml", "fetched_at": "2026-10-09T04:16:28.965Z", "raw_sha256": "01191bc9363c1e228945223307bfe15c5f89293e73c41891813d4e6526e4c94d" } }, "artificial-analysis-hle": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-hle", "name": "HLE · AA", "operator": "Artificial Analysis", "description": "横跨学科的专家级难题，不用工具作答。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Humanity’s Last Exam（Artificial Analysis 复跑）", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "what": "取 Artificial Analysis 智能指数中 Humanity’s Last Exam 的正确率（不用工具）。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的专家难题一类。", "limits": "部分题目答案存在争议；正确率普遍偏低，接近 0 的成绩只说明“至多这么强”。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.914Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-hle", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-hle", "fetched_at": "2026-10-09T04:16:29.244Z", "raw_sha256": "b5cedea42d535ea1a6c4a51de810f351621871b44036a361027ba612541f73f2" } }, "epoch-frontiermath": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3", "operator": "Epoch AI", "description": "研究级数学推理", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "FrontierMath v2 · Tiers 1–3", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "what": "只用 v2 的 Tiers 1–3；旧版题目和 Tier 4 不混入同一指标。", "usage": "计分：科研领域的研究级数学一类。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.635Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-frontiermath", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-frontiermath", "fetched_at": "2026-10-09T04:16:30.298Z", "raw_sha256": "a1af704aefe185a3253b0c5d37c7a901effa9538feab3e0de97459f1356e4486" } }, "livebench-reasoning": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学", "operator": "LiveBench", "description": "Reasoning + Mathematics 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 推理与数学", "area": "科研", "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "what": "读取 Reasoning 和 Mathematics 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：只计 Reasoning 一项（科研领域的逻辑一类）。Mathematics 头部已接近满分，暂不计分。", "limits": "两项均分衡量组合表现，不能当成 Reasoning 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-10-07T20:15:35.000Z", "syncedAt": "2026-10-09T00:05:24.537Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/livebench-reasoning", "api_url": "https://aihot.news/api/site/leaderboard/sources/livebench-reasoning", "fetched_at": "2026-10-09T04:16:30.441Z", "raw_sha256": "399a70109e6fcab8b25edd795c8a6fc3e919d26c04864552ffd374da9e650c8d" } }, "artificial-analysis-critpt": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-critpt", "name": "CritPt · AA", "operator": "Artificial Analysis", "description": "研究级物理问题，考验专业推导。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "CritPt（Artificial Analysis 复跑）", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "what": "取 Artificial Analysis 智能指数中 CritPt 的正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的研究级物理一类。", "limits": "题量较少，正确率普遍很低；题集正在接受公开复核。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.928Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-critpt", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-critpt", "fetched_at": "2026-10-09T04:16:30.448Z", "raw_sha256": "a7647288ba1294d2586e4648543bb735bfb88e532610f77cfd32076d5018b08f" } }, "epoch-chess": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "description": "根据文字棋局寻找正确走法", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "Chess Puzzles", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "what": "生成的棋局以 Stockfish 标准答案判分；输入为棋局文本，不归入视觉理解。", "usage": "计分：科研领域的棋类与游戏一类（与 Mystery Game Puzzles 合起来算一票）。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.667Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-chess", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-chess", "fetched_at": "2026-10-09T04:16:30.324Z", "raw_sha256": "2af49874eff7498cb851f5dd91190877a4ef92af7f18ebb12cd4e78311a680fe" } }, "epoch-mystery": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "description": "从陌生规则推导正确行动", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "Mystery Game Puzzles", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "what": "只取 Epoch 标准协议的 mean_score；题目不公开这一限制明确保留。", "usage": "计分：科研领域的棋类与游戏一类（与 Chess Puzzles 合起来算一票）。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.677Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-mystery", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-mystery", "fetched_at": "2026-10-09T04:16:31.523Z", "raw_sha256": "339e4f5d5bc142342d3ccc5721349ba8f0306c2413e7b7528c5544a7b026ba20" } }, "vals-mysterymechanism": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "description": "自己设计实验，去推断一套隐藏的规律。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MysteryMechanism（Vals AI）", "area": "科研", "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "what": "取 Vals 专项页 Overall 成绩。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的实验推断一类。", "limits": "参评模型较少；Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.860Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-mysterymechanism", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-mysterymechanism", "fetched_at": "2026-10-09T04:16:31.523Z", "raw_sha256": "e06502f24240ee9f603e9c0bca3a30da12b04f9212edcf9b724218f844c51c3b" } }, "vals-terminal-bench-science": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals", "operator": "Vals AI", "description": "在终端完成科学工作流与分析。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Terminal-Bench Science · Vals", "area": "科研", "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "what": "70 个专家审核的科学终端任务，统一 Mini-SWE-agent，由任务验证器检查结果。", "usage": "计分：科学工作流；与 AA 复跑的同一题库共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取版本 0.1 的 Mini-SWE-agent 成绩；Terminus 2 不混入。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-10-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.919Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-terminal-bench-science", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-terminal-bench-science", "fetched_at": "2026-10-09T04:16:32.184Z", "raw_sha256": "c965af80207391790a5404a75c3d760858482b9ebef423a1ebaebd7c395814f2" } }, "artificial-analysis-terminal-bench-science": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA", "operator": "Artificial Analysis", "description": "在终端完成科学工作流与分析。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Terminal-Bench Science · AA", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "what": "70 个 Harbour 科学终端任务，统一 mini-swe-agent，每题三次运行的平均 pass@1。", "usage": "计分：科学工作流；与 Vals 复跑的同一题库共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "取 terminalBenchScience 总体成绩，不混用领域切片。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.199Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-terminal-bench-science", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-terminal-bench-science", "fetched_at": "2026-10-09T04:16:31.590Z", "raw_sha256": "0bbd89bc926e2bacc664c6a88a8e2c3643b1225216a8d86432fe8cf276f279b7" } }, "epoch-frontiermath-tier4": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-frontiermath-tier4", "name": "FrontierMath v2 · Tier 4", "operator": "Epoch AI", "description": "更高难度的数学研究题", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "FrontierMath v2 · Tier 4", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tier-4-v2", "what": "独立保存 Tier 4，和 Tiers 1–3 共用数学证据预算；保留样本误差。", "usage": "Tier 4 题量太少，只作参考，不计分。", "limits": "Tier 4 题量太少，只作参考，不计分。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.654Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-frontiermath-tier4", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-frontiermath-tier4", "fetched_at": "2026-10-09T04:16:33.202Z", "raw_sha256": "3acf961026e764d487072ed3b00b88236524ef5d885b2128c78c0db7ba0501ca" } }, "hle": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "hle", "name": "Humanity’s Last Exam", "operator": "CAIS / Scale AI", "description": "横跨学科的高难度学术知识与推理。", "status": "observing", "brand": { "src": null, "monogram": "C", "raster": false }, "fullName": "Humanity’s Last Exam", "area": "知识", "officialUrl": "https://labs.scale.com/leaderboard/humanitys_last_exam", "what": "固定题集、工具条件和裁判；HLE、HLE-Rolling 与带工具结果分别比较。AA v4.3 已包含 HLE，将来若有可用专项成绩，只用于分类，不在综合榜重复计票。", "usage": "观察中：AA 统一复跑的 HLE 已计分；官方榜单接受各家自报成绩，暂不计分。", "limits": "观察中：AA 统一复跑的 HLE 已计分；官方榜单接受各家自报成绩，暂不计分。", "license": "评测代码 MIT；官方榜单成绩再展示边界尚待确认。", "attribution": "成绩由 CAIS / Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/hle", "api_url": "https://aihot.news/api/site/leaderboard/sources/hle", "fetched_at": "2026-10-09T04:16:33.260Z", "raw_sha256": "fb4a760d71ce26e3982b4ca3efe1d34dab7855b0f02a729d26ebc3e099b1d264" } }, "simplebench": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "simplebench", "name": "SimpleBench", "operator": "SimpleBench", "description": "用日常常识和逻辑问题，检验模型能否识别问题中的实际约束。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SimpleBench", "area": "科研", "officialUrl": "https://simple-bench.com/", "what": "区分选择题与开放回答，固定 AVG@5、温度和提示协议；不把加入榜单日期当逐题评测时间。", "usage": "候选观察：新品覆盖及时，等待官网成绩复用边界与可冻结的数据版本说明；不套用代码许可，尚未自动采集或计分。", "limits": "日常常识题具有补充价值，但题目代表性与选择题/开放回答条件均影响解释。", "license": "样题与评测代码仓库 MIT；官网独立发布的最新成绩脚本再展示边界尚未确认。", "attribution": "成绩由 SimpleBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/simplebench", "api_url": "https://aihot.news/api/site/leaderboard/sources/simplebench", "fetched_at": "2026-10-09T04:16:33.411Z", "raw_sha256": "834b742060314c8e44023b379e06c099b663077ec12a44697ec6f9dc5039b0ed" } }, "matharena-lean": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "matharena-lean", "name": "MathArena · ArXivLean", "operator": "MathArena / ETH Zurich", "description": "用 Lean 验证数学证明", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · ArXivLean", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/matharena-lean", "api_url": "https://aihot.news/api/site/leaderboard/sources/matharena-lean", "fetched_at": "2026-10-09T04:16:33.239Z", "raw_sha256": "261ddf9a06786909e277e72b20a4cd2da3cc580c7feb6e7e88ee2bded294ad71" } }, "matharena-broken": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "matharena-broken", "name": "MathArena · BrokenArXiv", "operator": "MathArena / ETH Zurich", "description": "发现和修复数学证明中的错误", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · BrokenArXiv", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/matharena-broken", "api_url": "https://aihot.news/api/site/leaderboard/sources/matharena-broken", "fetched_at": "2026-10-09T04:16:33.830Z", "raw_sha256": "716b30fc58aaee5e3dd7bf93173b48a6637dbf006bca14168754804a5b657e07" } }, "matharena-math": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "matharena-math", "name": "MathArena · ArXivMath", "operator": "MathArena / ETH Zurich", "description": "从近期数学论文提炼的新题", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · ArXivMath", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/matharena-math", "api_url": "https://aihot.news/api/site/leaderboard/sources/matharena-math", "fetched_at": "2026-10-09T04:16:34.625Z", "raw_sha256": "8d96ab3d7beede3c965d53b16fdbf803d5ad448b842648255f1e1603f9993b74" } }, "arc-agi-2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arc-agi-2", "name": "ARC-AGI-2", "operator": "ARC Prize", "description": "从陌生规则中学习和泛化", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "ARC-AGI-2", "area": "科研", "officialUrl": "https://arcprize.org/leaderboard", "what": "固定标准 harness，与提供商专门优化的 adapter 成绩分开。", "usage": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "limits": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 ARC Prize 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arc-agi-2", "api_url": "https://aihot.news/api/site/leaderboard/sources/arc-agi-2", "fetched_at": "2026-10-09T04:16:34.624Z", "raw_sha256": "dffb8696de48377376b638fe85053dd02a5d016a84efc415a0285de84dd1315b" } }, "arc-agi-3": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arc-agi-3", "name": "ARC-AGI-3", "operator": "ARC Prize", "description": "从陌生规则中学习和泛化", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "ARC-AGI-3", "area": "科研", "officialUrl": "https://arcprize.org/leaderboard", "what": "固定标准 harness，与提供商专门优化的 adapter 成绩分开。", "usage": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "limits": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 ARC Prize 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "reasoning", "name": "科研" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arc-agi-3", "api_url": "https://aihot.news/api/site/leaderboard/sources/arc-agi-3", "fetched_at": "2026-10-09T04:16:33.951Z", "raw_sha256": "9c887145ea500196e13bf1233a9f3cb5ad10ee584eb48ceb9fa8e38d3c227f0c" } }, "artificial-analysis-gdpval": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "description": "按真实职业交付件出题，两两比较产出质量。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "GDPval-AA（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "what": "取 Artificial Analysis 智能指数中 GDPval-AA 的 Elo 等级分。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域，与 AA-Briefcase 同属职业交付一类，两项合起来算一票。", "limits": "等级分来自模型裁判两两比较，不是正确率；它的刻度按同一批模型在其他领域的水平对齐，不能自定尺度。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.941Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-gdpval", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-gdpval", "fetched_at": "2026-10-09T04:16:36.090Z", "raw_sha256": "3495d6a23ca6755784e7f89d0085ef9ed027b926d4d783d7dde2128e005dd796" } }, "artificial-analysis-briefcase": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "description": "办公知识工作的长任务，两两比较交付结果。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Briefcase（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "what": "取 Artificial Analysis 智能指数中 AA-Briefcase 的 Elo 等级分。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域，与 GDPval-AA 同属职业交付一类，两项合起来算一票。", "limits": "等级分来自模型裁判两两比较，不是正确率；它的刻度按同一批模型在其他领域的水平对齐，不能自定尺度。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.016Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-briefcase", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-briefcase", "fetched_at": "2026-10-09T04:16:36.387Z", "raw_sha256": "db67815e95c00eacf982af1fc419448530c4d3a9e31333b69b02e94b3f71ecf6" } }, "artificial-analysis-automationbench": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-automationbench", "name": "AutomationBench · AA", "operator": "Artificial Analysis", "description": "跨办公软件完成一整条业务自动化流程。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AutomationBench-AA（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "what": "取 Artificial Analysis 智能指数中 AutomationBench-AA 的完成率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的业务流程一类。", "limits": "在 AA 的统一环境中运行，与 Zapier 官方榜单的私有测试集不同。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.084Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-automationbench", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-automationbench", "fetched_at": "2026-10-09T04:16:35.137Z", "raw_sha256": "2a61a6bffa5e435f93d0936b7525eb25188f155440cbd4a800865de82b9ebe44" } }, "vals-finance-agent": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-finance-agent", "name": "Vals Finance Agent", "operator": "Vals AI", "description": "金融分析师的日常办公题，看研报、建模这类工作做得怎样。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vals Finance Agent", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "what": "取 Vals 专项页 Overall 正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的金融一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.756Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-finance-agent", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-finance-agent", "fetched_at": "2026-10-09T04:16:35.806Z", "raw_sha256": "01603d09568388fb0dc82a5535997588dd074a79857b0c74ec2257ddc67f0eb0" } }, "mercor-apex-agents": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1", "operator": "Mercor", "description": "投行、咨询和法律里的长任务，看模型当助手能不能把一件完整的工作做完。", "status": "ranked", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "fullName": "Mercor APEX-Agents 1.1", "area": "专业办公", "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "what": "固定 1.1 数据集和官方 Loop 环境，采用 Pass@1 及其对应误差，按预先固定的推理档位选配置。Mean Score 是评分项平均完成度，不混作任务完成率。", "usage": "计分：专业办公领域的投行、咨询与律所长任务一类。", "limits": "限投行、咨询、法律任务；原页未提供逐型号评测日期，少量运行缺失的原因尚未公开。不能将此分数解释为所有办公工作的可靠性。", "license": "官方公开结果；数据与代码许可不等于榜单再分发授权", "attribution": "成绩由 Mercor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-09-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:29.879Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/mercor-apex-agents", "api_url": "https://aihot.news/api/site/leaderboard/sources/mercor-apex-agents", "fetched_at": "2026-10-09T04:16:37.467Z", "raw_sha256": "d8c9d315f2c7df2e9fffe9421b5de5b407b031be0f58746222fe8e258ac5e002" } }, "vals-harvey-lab": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-harvey-lab", "name": "Harvey Legal Agent Benchmark", "operator": "Vals AI / Harvey", "description": "处理法律资料并交付 Word、Excel、幻灯片等可审阅文件。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Harvey Legal Agent Benchmark", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/hlab", "what": "固定 Harvey 私有测试集、文件工具和 docx、xlsx、pptx 技能，禁用互联网；取两名裁判的整任务通过率均值。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的法律一类。", "limits": "整项任务通过率普遍很低；Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI / Harvey 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-10-07T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.854Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-harvey-lab", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-harvey-lab", "fetched_at": "2026-10-09T04:16:36.979Z", "raw_sha256": "5c96748c239a1a5270c2a3cc39f93f269500c3549bb7d95899051d55a820709d" } }, "mercor-tau3-banking": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor", "operator": "Mercor", "description": "在统一工具环境里办理银行客服业务：查政策、找产品、改账户。", "status": "ranked", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "fullName": "τ³-Banking（Mercor 复跑）", "area": "专业办公", "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "what": "取 Mercor 给所有模型统一复跑的 pass@1。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的客服一类。", "limits": "依赖模拟用户与工具协议，只代表银行客服场景；Sierra 官方榜单接受厂商自报成绩，不计分。", "license": "官方公开结果；题目与评测框架来自 Sierra tau2-bench（MIT）", "attribution": "成绩由 Mercor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:29.891Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/mercor-tau3-banking", "api_url": "https://aihot.news/api/site/leaderboard/sources/mercor-tau3-banking", "fetched_at": "2026-10-09T04:16:37.514Z", "raw_sha256": "487a9a82bf3e9d7362200df34068102f121b4ee541f6cc7222d480d7ebfb1fbc" } }, "livebench-data-analysis": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析", "operator": "LiveBench", "description": "表格重排、列类型标注、表连接预测等数据处理任务。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 数据分析", "area": "专业办公", "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "what": "读取 Data Analysis 分类的未舍入成绩。", "usage": "计分：专业办公领域的表格数据处理一类。", "limits": "题目每期更新，格式要求严格；组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "成绩由 LiveBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-10-07T20:15:35.000Z", "syncedAt": "2026-10-09T00:05:24.544Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/livebench-data-analysis", "api_url": "https://aihot.news/api/site/leaderboard/sources/livebench-data-analysis", "fetched_at": "2026-10-09T04:16:37.513Z", "raw_sha256": "b0e9b3f2cf328fe23f65727c933ef652a35c96d284603677e98180618e173a2b" } }, "vals-medcode": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "description": "根据住院记录给出医疗诊断编码。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MedCode", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/medcode", "what": "去标识化住院记录，由两名认证编码员标注并解决分歧；统一温度 1、最多 30k 输出 token，核对 ICD-10-CM 主次诊断编码。", "usage": "计分：医疗编码一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "取总体编码得分，不等同于整份病例完全正确率；不混用主次诊断切片。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-10-08T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:42.939Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-medcode", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-medcode", "fetched_at": "2026-10-09T04:16:37.958Z", "raw_sha256": "3481be7ec71c5f6e46fb8ec17dd8a847ddcf16de6f88a0e936879b26238e1234" } }, "artificial-analysis-analyst-agent": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "description": "可靠地分析数据并回答专业问题。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Analyst Agent", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "what": "80 个私有问题、14 个领域，统一 Stirrup 工具环境；每题独立运行五次，五次全部成功才通过（pass^5）。", "usage": "计分：数据分析；与 LiveBench Data Analysis 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "pass^5 不是五次至少成功一次的 pass@5。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.154Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-analyst-agent", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-analyst-agent", "fetched_at": "2026-10-09T04:16:42.293Z", "raw_sha256": "54faff275af295bd61f983ebfa662a220be25a917a07f4dee4b7cf3415088a58" } }, "tau-banking": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "tau-banking", "name": "τ³-Banking", "operator": "Sierra", "description": "在统一工具环境中办理银行业务，检验规则理解、客户沟通与任务完成。", "status": "observing", "brand": { "src": "/leaderboard-sources/sierra.png", "monogram": "S", "raster": true }, "fullName": "τ³-Banking", "area": "专业办公", "officialUrl": "https://taubench.com/leaderboard/?benchmark=knowledge", "what": "只比较官方 standard 银行业务、v1.0.1、alltools 和 GPT-5.2 用户模拟器；采用 pass@1。不同工具、提示、模拟器、语音模式与自定义系统不混排。", "usage": "观察中：Mercor 统一复跑的 τ³-Banking 已计分；官方榜单接受厂商自报成绩，不计分。", "limits": "成绩属于固定银行任务系统，依赖模拟用户与工具协议；不是所有职业的通用正确率。", "license": "MIT · Sierra Research tau2-bench 官方仓库；保留来源署名。", "attribution": "成绩由 Sierra 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/tau-banking", "api_url": "https://aihot.news/api/site/leaderboard/sources/tau-banking", "fetched_at": "2026-10-09T04:16:38.652Z", "raw_sha256": "d0f57bb48f18b0c2edadbfdd5a60bf34455f4f2055d952485e6836ffbad263d3" } }, "zapier-automation": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "zapier-automation", "name": "AutomationBench", "operator": "Zapier", "description": "跨办公软件完成自动化工作", "status": "observing", "brand": { "src": null, "monogram": "Z", "raster": false }, "fullName": "AutomationBench", "area": "专业办公", "officialUrl": "https://zapier.com/benchmarks", "what": "只比较同一官方私有测试集的配置；Fable 5.1 混用 Opus 5 的回退行必须排除。", "usage": "观察中：AA 统一复跑的 AutomationBench 已计分；官方私有排行榜数据暂不计分。", "limits": "观察中：AA 统一复跑的 AutomationBench 已计分；官方私有排行榜数据暂不计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Zapier 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/zapier-automation", "api_url": "https://aihot.news/api/site/leaderboard/sources/zapier-automation", "fetched_at": "2026-10-09T04:16:38.090Z", "raw_sha256": "e2367467ffb2f661e60c3e06276c7c655a7a021867cd1d14c577254305d2584b" } }, "officeqa-pro": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "officeqa-pro", "name": "OfficeQA Pro", "operator": "Databricks", "description": "从大量真实财务文件中检索、计算并回答问题。", "status": "observing", "brand": { "src": null, "monogram": "D", "raster": false }, "fullName": "OfficeQA Pro", "area": "专业办公", "officialUrl": "https://github.com/databricks/officeqa", "what": "Pro 固定 133 题，与 Full 246 题及 Pro V2 的 90 题分开；端到端检索与直接提供答案所在页面属于不同条件。Claude Agent SDK、Codex SDK、Antigravity CLI 的成绩属于各自运行系统，不能混作统一基础模型评测。它衡量文件理解，不代表办公文件的产出质量。", "usage": "观察中：官方继续更新，但目前主要比较不同产品运行系统，原始汇总协议仍待核实；保留资料理解参考，不将不同工具系统的成绩直接归给基础模型。", "limits": "代码及题库许可不自动覆盖所有外部成绩；新旧题集、端到端与答案页面直供条件不能混排，当前没有自动采集成绩。", "license": "代码 Apache 2.0；题目数据 CC BY-SA 4.0，现迁至需要申请访问的 Hugging Face 数据集。", "attribution": "成绩由 Databricks 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/officeqa-pro", "api_url": "https://aihot.news/api/site/leaderboard/sources/officeqa-pro", "fetched_at": "2026-10-09T04:16:43.760Z", "raw_sha256": "4602aac2d12a650363ccfa0e75800d3a383d27974a903fde5ae0a927ce41ef72" } }, "mcp-atlas": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "mcp-atlas", "name": "MCP Atlas", "operator": "Scale AI", "description": "通过真实 MCP 服务检索资料、操作文档与数据库，完成跨软件任务。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "MCP Atlas", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/mcp_atlas", "what": "固定 2026 年 4 月调整后的运行协议和 100 次工具调用预算；完整 1,000 题与公开 500 题分开。只取任务通过率，不把回答条件覆盖率重复计票；裁判、重试和工具版本需随成绩批次锁定。", "usage": "观察中：可补跨办公软件的工具执行证据；待确认稳定成绩导出、实际评测批次和再展示边界，目前只收录来源，不自动采集或计分。", "limits": "网页旧方法说明和实时成绩可能不同步；公开题集代码不是实时成绩导出。与 Scale 的其他榜共属同一机构，不增加独立机构数。", "license": "官方环境代码采用 MIT；该许可不自动覆盖实时网页成绩，榜单数据再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/mcp-atlas", "api_url": "https://aihot.news/api/site/leaderboard/sources/mcp-atlas", "fetched_at": "2026-10-09T04:16:42.834Z", "raw_sha256": "584dd6e616a65a5757266c0b56678220a6c570b297ddc03102c10fda49a8bb52" } }, "financebenchmark": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "financebenchmark", "name": "FinanceBenchmark", "operator": "FinanceBenchmark", "description": "完成金融定价、资本计算和风控任务，用数值与代码执行结果核验。", "status": "observing", "brand": { "src": null, "monogram": "F", "raster": false }, "fullName": "FinanceBenchmark", "area": "专业办公", "officialUrl": "https://finbenchmark.ai/", "what": "题集 v2、v3 与运行环境版本分开；保留每题三次尝试的条件。官网 Tasks passed 指至少一次成功，却同时称为 Pass@1，不能与单次平均正确率混用；Finance Index 也是另一个汇总指标。", "usage": "观察中：可补确定答案与代码验证的金融证据；指标口径仍需核对，官方 Sol 结果 JSON 下载本次返回 404，当前只收录来源，不自动采集或计分。", "limits": "三次至少一次成功不等于单次成功率；区间必须对应正确指标，滚动模型别名也需核验版本。题库和下载按钮公开不等于原始成绩已能获取。", "license": "官方仓库代码为 MIT，网站要求保留 FinanceBenchmark 引用；未确认实时成绩完整导出的使用边界，不将代码许可替代榜单数据许可。", "attribution": "成绩由 FinanceBenchmark 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/financebenchmark", "api_url": "https://aihot.news/api/site/leaderboard/sources/financebenchmark", "fetched_at": "2026-10-09T04:16:43.457Z", "raw_sha256": "a1bc8d5faca2ae3ad2a6b9931aae296b611a3ccc42d96e32714880c2e195ed41" } }, "prbench-legal": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "prbench-legal", "name": "PRBench · Legal", "operator": "Scale AI", "description": "回答真实法律工作中的复杂问题，检验专业判断和论证质量。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "PRBench · Legal", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/prbench-legal", "what": "固定 Legal 完整 500 题或 Hard 250 题之一；搜索工具、推理档位和裁判版本保持一致。采用该题集的 rubric 汇总分，不把完整集与困难子集重复计票，也不将专业问答描述为完整文件交付。", "usage": "观察中：可补法律专业判断；先核实题集、工具条件和成绩批次，目前只收录来源，不自动采集或计分。", "limits": "由模型裁判评审，存在裁判偏好；同机构的法律和金融专项应共用影响边界，不能作为两家独立证据。它不直接衡量 Word、Excel 或幻灯片的制作能力。", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/prbench-legal", "api_url": "https://aihot.news/api/site/leaderboard/sources/prbench-legal", "fetched_at": "2026-10-09T04:16:43.994Z", "raw_sha256": "cd7d77c12080707fb902247b790cca043e2edd8595f90e105d19689ca923c109" } }, "prbench-finance": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "prbench-finance", "name": "PRBench · Finance", "operator": "Scale AI", "description": "回答真实金融决策问题，检验专业判断、推导和风险说明。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "PRBench · Finance", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/prbench-finance", "what": "固定 Finance 完整 600 题或 Hard 300 题之一；搜索工具、推理档位和裁判版本保持一致。采用该题集的 rubric 汇总分，不将完整集和困难子集重复计票，也不把专业问答当作端到端办公任务。", "usage": "观察中：可补金融专业判断；待核实题集、工具条件、稳定导出和再展示边界，目前只收录来源，不自动采集或计分。", "limits": "由模型裁判评审，不代表已经完成财务建模或表格制作；与 PRBench Legal、MCP Atlas 同属 Scale，不能靠增加榜单数量提高机构权重。", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/prbench-finance", "api_url": "https://aihot.news/api/site/leaderboard/sources/prbench-finance", "fetched_at": "2026-10-09T04:16:44.446Z", "raw_sha256": "53302ba93218792f8672915f95f3abdb30911eed8946d76e0cafae9c41a18681" } }, "spreadsheetbench-2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "spreadsheetbench-2", "name": "SpreadsheetBench 2", "operator": "SpreadsheetBench / Renmin University / AfterQuery", "description": "完成财务建模、修复工作簿和数据可视化的整套表格工作。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SpreadsheetBench 2", "area": "专业办公", "officialUrl": "https://spreadsheetbench.github.io/", "what": "固定 V2 的 321 个任务和统一 SWE-agent 条件；独立办公产品和其他 scaffold 成绩不归给基础模型。表格可视化的正确性不等于广泛审美。", "usage": "观察中：已有稳定 JSON 和逐行日期，但当前统一条件下的基础模型覆盖不足；等待当期型号实测后再接入分数。", "limits": "观察中：已有稳定 JSON 和逐行日期，但当前统一条件下的基础模型覆盖不足；等待当期型号实测后再接入分数。", "license": "官方网站页尾声明 MIT；数据集与代码入口公开。", "attribution": "成绩由 SpreadsheetBench / Renmin University / AfterQuery 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/spreadsheetbench-2", "api_url": "https://aihot.news/api/site/leaderboard/sources/spreadsheetbench-2", "fetched_at": "2026-10-09T04:16:45.863Z", "raw_sha256": "bbdef4264e8a7df75781952c0d79c714b453ff8a4c736e5183d6b8cdf4b96a2a" } }, "vending-bench-2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vending-bench-2", "name": "Vending-Bench 2", "operator": "Andon Labs", "description": "经营模拟商店一年，处理采购、谈判、库存与定价。", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "Vending-Bench 2", "area": "专业办公", "officialUrl": "https://andonlabs.com/evals/vending-bench-2", "what": "比较同一 V2 单代理任务的期末余额，保留多次运行方差；多人竞争 Arena、不同工具和推理配置分别处理。任务中的经营成果是评测成绩，API 价格不计能力分。", "usage": "观察中：补充长期经营与电商能力的价值明确；稳定原始成绩协议、运行配置与再展示许可仍待补齐。盈利也可能来自利用模拟器，不能直接等同可信经营。", "limits": "观察中：补充长期经营与电商能力的价值明确；稳定原始成绩协议、运行配置与再展示许可仍待补齐。盈利也可能来自利用模拟器，不能直接等同可信经营。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Andon Labs 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vending-bench-2", "api_url": "https://aihot.news/api/site/leaderboard/sources/vending-bench-2", "fetched_at": "2026-10-09T04:16:45.107Z", "raw_sha256": "0956e5cddea195db5c4636b3ac9f6ec6548280cc9d6a38d4ba9895d2d734ccc1" } }, "epoch-ebr": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-ebr", "name": "EBR-bench", "operator": "Epoch AI", "description": "在长任务中学习新规则并使用记忆", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "EBR-bench", "area": "专业办公", "officialUrl": "https://epoch.ai/benchmarks/ebr-bench", "what": "公开汇总混合了记忆压缩配置；未取得固定配置的可比成绩前仅供参考。", "usage": "仅供参考，不进入综合或分类分数。", "limits": "仅供参考，不进入综合或分类分数。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": "2026-09-29T18:13:25.000Z", "syncedAt": "2026-10-09T00:05:26.716Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-ebr", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-ebr", "fetched_at": "2026-10-09T04:16:45.064Z", "raw_sha256": "2e3f2109d9a943974751f5cd09407ea1cd97e348bb53ac3c2b191ac5777627ad" } }, "vals-excel": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-excel", "name": "Excel · EMB", "operator": "Vals AI", "description": "表格编辑与 Excel 办公", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Excel · EMB", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/emb", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-excel", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-excel", "fetched_at": "2026-10-09T04:16:46.982Z", "raw_sha256": "4359113980a928dca952587534033c5a46b0ca0c6bd1832980f08bc617b018ae" } }, "vals-legal": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-legal", "name": "Legal Research", "operator": "Vals AI", "description": "法律检索和论证", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Legal Research", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/legal_research", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-legal", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-legal", "fetched_at": "2026-10-09T04:16:46.422Z", "raw_sha256": "de247e4db9c0b8d8765bafc5d33ed52fef0b648d53fc560f9949d24aa457c002" } }, "vals-tax": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-tax", "name": "TaxAgentBench", "operator": "Vals AI", "description": "税务专业问题", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "TaxAgentBench", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/tax_agent_bench", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-tax", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-tax", "fetched_at": "2026-10-09T04:16:47.060Z", "raw_sha256": "347f32534db662d76df469eb7d6ee02d13791a69ccb9a984e2ba471575626e07" } }, "vals-medscribe": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-medscribe", "name": "MedScribe", "operator": "Vals AI", "description": "临床记录整理与撰写", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MedScribe", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/medscribe", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-medscribe", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-medscribe", "fetched_at": "2026-10-09T04:16:46.988Z", "raw_sha256": "32a4a763ed30d6e82772f56a56fdec515195f05f9dbec0a9bef1bf4bdf8e680f" } }, "vals-biomystery": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "vals-biomystery", "name": "BioMysteryBench", "operator": "Vals AI", "description": "生物学研究推理", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "BioMysteryBench", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/biomysterybench", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/vals-biomystery", "api_url": "https://aihot.news/api/site/leaderboard/sources/vals-biomystery", "fetched_at": "2026-10-09T04:16:48.806Z", "raw_sha256": "f3117819d8e736201ae52d702376d42a828f3e42075298c8c9aebe26b70148a1" } }, "terminal-science": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "terminal-science", "name": "Terminal-Bench Science", "operator": "Harbor", "description": "在终端里完成科研任务", "status": "observing", "brand": { "src": null, "monogram": "H", "raster": false }, "fullName": "Terminal-Bench Science", "area": "专业办公", "officialUrl": "https://www.terminal-bench-science.ai/", "what": "区分 Claude Code、Codex 和 mini-SWE 等系统；不能把不同系统成绩直接当模型能力。", "usage": "观察中：12 个型号且混合运行系统，新前沿型号仍不完整。", "limits": "观察中：12 个型号且混合运行系统，新前沿型号仍不完整。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Harbor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "professional", "name": "专业办公" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/terminal-science", "api_url": "https://aihot.news/api/site/leaderboard/sources/terminal-science", "fetched_at": "2026-10-09T04:16:48.778Z", "raw_sha256": "3ad50f5d1cbe2186afe726e19a702dda2d2d819ef36afb80aec711333d98ed97" } }, "artificial-analysis-omniscience": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "description": "事实问答，答错倒扣、拒答不扣，看模型知道什么、会不会乱编。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Omniscience（Artificial Analysis）", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "what": "取 Artificial Analysis 智能指数中 Omniscience 的分数（−100 到 100）。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域，与 SimpleQA Verified 同属事实准确一类，两项合起来算一票。", "limits": "分数同时反映知识量和是否乱编，不单纯是知识量。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.106Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-omniscience", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-omniscience", "fetched_at": "2026-10-09T04:16:48.239Z", "raw_sha256": "fd3712c89bcef3eab63f79bb45dce9233f63760d1ec89f19cebba75912482aa5" } }, "epoch-simpleqa": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "description": "短问题的事实准确性", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "SimpleQA Verified", "area": "知识", "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "what": "只取 Epoch 自行复跑的 mean_score，采用固定推理档位；不读取 ECI 或 Best score 聚合列。", "usage": "计分：知识问答领域，与 AA-Omniscience 同属事实准确一类，两项合起来算一票。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.685Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-simpleqa", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-simpleqa", "fetched_at": "2026-10-09T04:16:48.291Z", "raw_sha256": "b546145b0db24448a4f55a12dc47f550f54ba146fc7ce4a333012822ae9406bf" } }, "artificial-analysis-lcr": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "description": "读很长的文档后，回答需要综合推理的问题。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-LCR（Artificial Analysis）", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "what": "取 Artificial Analysis 智能指数中 AA-LCR 的正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域的长文理解一类。", "limits": "只考英文长文。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.132Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-lcr", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-lcr", "fetched_at": "2026-10-09T04:16:49.926Z", "raw_sha256": "8133491658d9c61dec11e53f952a78e2b994ecdfed46451efe25a00126cecd26" } }, "livebench-writing": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "livebench-writing", "name": "LiveBench · 语言与指令", "operator": "LiveBench", "description": "Language + IF 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 语言与指令", "area": null, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "what": "读取 Language 和 IF 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：Language 与 IF 分别计入知识问答领域的语言理解、按要求作答两类。", "limits": "两项均分衡量组合表现，不能当成 Language 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": "2026-10-07T20:15:35.000Z", "syncedAt": "2026-10-09T00:05:24.517Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/livebench-writing", "api_url": "https://aihot.news/api/site/leaderboard/sources/livebench-writing", "fetched_at": "2026-10-09T04:16:49.879Z", "raw_sha256": "39b967da193b0a2ce3107bb44e8a4774c708193f853a2d972978da0b395fb07d" } }, "bullshitbench-v2": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提", "operator": "Peter Gostev / BullshitBench", "description": "识别问题里的错误前提，观察模型会不会顺着错误继续编造。", "status": "ranked", "brand": { "src": null, "monogram": "P", "raster": false }, "fullName": "BullshitBench v2", "area": "知识", "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "what": "取官方 v2 最新榜单的 green rate：所有尝试中明确识破错误前提的比例。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域的识破错误前提一类。", "limits": "全部为错误前提题，不代表一般知识广度；由模型裁判判定。", "license": "MIT · Copyright (c) 2026 Peter Gostev；官方成绩随同仓库发布，无单独数据许可例外。", "attribution": "成绩由 Peter Gostev / BullshitBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": "2026-09-29T20:50:19.000Z", "syncedAt": "2026-10-09T00:05:44.778Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/bullshitbench-v2", "api_url": "https://aihot.news/api/site/leaderboard/sources/bullshitbench-v2", "fetched_at": "2026-10-09T04:16:50.612Z", "raw_sha256": "e9661ec72f5e6f2ecd007e41dc0de64bf1483707dc1c955de92e549fe9c154ed" } }, "artificial-analysis-mlcr": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "description": "准确、完整且简洁地理解医疗长记录。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "MLCR-AA", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "what": "25k–64k token 医疗记录；专家参考答案、三个模型裁判多数表决；每题三次，取总体 pass@1，空答或超过参考答案五倍长度失败。", "usage": "计分：长文理解；与 AA-LCR 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 mlcrOverall，包含长度失败；不替换成仅对合格长度作答的准确率。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:21.165Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-mlcr", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-mlcr", "fetched_at": "2026-10-09T04:16:49.870Z", "raw_sha256": "4256a26af7c919261efebf8f41fda07234392852f8108ea79040a5111c9815c9" } }, "epoch-gpqa": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "epoch-gpqa", "name": "GPQA Diamond", "operator": "Epoch AI", "description": "研究生水平的物理、化学与生物知识", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "GPQA Diamond", "area": "知识", "officialUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "what": "只取 Epoch 自行复跑的 mean_score，保留推理档位与标准误。198 道科学知识题需要推理，但不能代表全部世界知识。", "usage": "头部模型已接近满分，区分度不足，只作参考，不计分。", "limits": "头部模型已接近满分，区分度不足。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": "2026-09-29T22:00:34.000Z", "syncedAt": "2026-10-09T00:05:26.703Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/epoch-gpqa", "api_url": "https://aihot.news/api/site/leaderboard/sources/epoch-gpqa", "fetched_at": "2026-10-09T04:16:51.686Z", "raw_sha256": "4c556cee3478b2b82715ba7bd2011cc5ffc8690b7b7cabb1eb236727842ee54a" } }, "facts-parametric": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "facts-parametric", "name": "FACTS Parametric", "operator": "Google DeepMind / Kaggle", "description": "不借助搜索工具回答世界事实问题，检验模型自身知识的准确性。", "status": "observing", "brand": { "src": null, "monogram": "G", "raster": false }, "fullName": "FACTS Parametric", "area": "知识", "officialUrl": "https://www.kaggle.com/benchmarks/google/facts-parametric/leaderboard", "what": "只考虑固定任务版本的 Score，保留公开集、私有集与置信区间供核查；不能把三个相关分数重复计票。FACTS 不在 AA v4.3 的十项底层评测中。", "usage": "观察中：官方协议与许可已确认，但没有完成题数或完整性标记；部分行区间显著宽，总分与公开集、私有集的汇总口径也存在待确认情形。等待评测方说明，不删除异常行来强行接入，不占评分预算。", "limits": "裁判与评测发起方均来自 Google，需结合其他机构证据。导出未说明部分结果是否完成同一套题，不能把局部结果当完整评测；从区间反推的样本量不是官方确认事实。", "license": "Apache 2.0 · 官方 benchmark 与 Score task 页面明确声明；Kaggle 文档提供公共榜单的免登录 JSON 下载。", "attribution": "成绩由 Google DeepMind / Kaggle 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "knowledge", "name": "知识问答" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/facts-parametric", "api_url": "https://aihot.news/api/site/leaderboard/sources/facts-parametric", "fetched_at": "2026-10-09T04:16:51.158Z", "raw_sha256": "e0957458922522a107edb7ea4253985c58dd7eae7397789e25847755dd1594d7" } }, "arena-text": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arena-text", "name": "Arena Text", "operator": "LMArena", "description": "真人匿名两两比较回答，看大家更愿意读哪一个。补上选择题看不出的整体体验。", "status": "ranked", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Text Style-Controlled", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "只取官方发布快照中的 Text Overall style-controlled 主榜，不抓取 arena.ai 实时页面。", "usage": "计分：真人盲选领域，只这一项。", "limits": "人类偏好受回答风格、用户结构和提示分布影响。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "general", "name": "真人盲选与综合指数" } }, "upstreamAt": "2026-10-02T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:21.217Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arena-text", "api_url": "https://aihot.news/api/site/leaderboard/sources/arena-text", "fetched_at": "2026-10-09T04:16:51.113Z", "raw_sha256": "f90033508e1c27cd414c82c9299cb60fe4f4ad7689bd0d2abf965626245ae030" } }, "artificial-analysis": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis", "name": "AA Index", "operator": "Artificial Analysis", "description": "把多项当前评测放到同一套环境里复跑，得到一个综合能力分。价格只作参考，不进总分。", "status": "cross_reference", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Artificial Analysis Intelligence Index", "area": null, "officialUrl": "https://artificialanalysis.ai/", "what": "取 Artificial Analysis 公开页面上的智能指数，只作交叉核对。", "usage": "综合指数不计分，只用来核对；其中 9 项分项已分别计分，不重复计入。", "limits": "聚合指数包含多种任务和运行条件；须保留指数版本、配置与上游署名。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "general", "name": "真人盲选与综合指数" } }, "upstreamAt": null, "syncedAt": "2026-10-09T00:05:20.872Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis", "fetched_at": "2026-10-09T04:16:51.858Z", "raw_sha256": "520fc69ce5a7ea1481fed5b78b32a8bdffceee847dd8172adf7dfbc325c70ff3" } }, "livebench-general": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "livebench-general", "name": "LiveBench", "operator": "LiveBench", "description": "题库持续更换，答案可以自动核对。用来看模型换题之后还能不能站住。", "status": "cross_reference", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench Global Average", "area": null, "officialUrl": "https://livebench.ai/", "what": "保留官方 Global Average 作整体参考；正式分数使用互不重复的专项任务。", "usage": "LiveBench 总均分只做交叉核对，不计分；它的分类成绩已分别计分。", "limits": "综合均分会稀释特定能力的变化；仅交叉参考，不重复计票。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "成绩由 LiveBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "general", "name": "真人盲选与综合指数" } }, "upstreamAt": "2026-10-07T20:15:35.000Z", "syncedAt": "2026-10-09T00:05:24.509Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/livebench-general", "api_url": "https://aihot.news/api/site/leaderboard/sources/livebench-general", "fetched_at": "2026-10-09T04:16:52.925Z", "raw_sha256": "453247a04756acd8d02d8715631cfe8b135c71aa94597267aa13255f58bb9260" } }, "arena-webdev": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arena-webdev", "name": "Arena WebDev", "operator": "LMArena", "description": "同样是真人盲选，比的是网页好不好用、能不能交付。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena WebDev", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "读取官方发布的 WebDev / Code Arena 快照，保留上游型号与配置。", "usage": "只作参考，不计分：部分型号经厂商自己的 Agent（如 codex-harness）参赛。", "limits": "不能把不同专用 Agent 或不同执行流程的交付表现直接归因于底层型号；待共同协议范围核实后再评估。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "general", "name": "真人盲选与综合指数" } }, "upstreamAt": "2026-10-06T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:21.244Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arena-webdev", "api_url": "https://aihot.news/api/site/leaderboard/sources/arena-webdev", "fetched_at": "2026-10-09T04:16:52.388Z", "raw_sha256": "e29708dbbabbc861a97303c6650e920ebe5f5df386839d716389271129d96b52" } }, "arena-creative-writing": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arena-creative-writing", "name": "Arena 创作盲选", "operator": "LMArena", "description": "真人盲选故事、诗歌和其他创意表达，观察作品是否吸引读者。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Creative Writing", "area": null, "officialUrl": "https://arena.ai/leaderboard/text/creative-writing", "what": "读取官方数据集 text_style_control 的 creative_writing 分类，采用风格控制后的 Bradley–Terry 分与置信区间；不借用整体文本或职业写作分类成绩。", "usage": "写作目前不计分，只作参考。", "limits": "写作目前不计分，只作参考。", "license": "CC BY 4.0 · Arena 官方 Hugging Face leaderboard-dataset；保留署名、来源链接并说明聚合改动。", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": "2026-10-02T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:21.235Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arena-creative-writing", "api_url": "https://aihot.news/api/site/leaderboard/sources/arena-creative-writing", "fetched_at": "2026-10-09T04:16:53.002Z", "raw_sha256": "33e844a55def68fcf5a407cfd7bc4f8f4cdcd145f3c1de1408f0973a7249d5bb" } }, "eq-creative": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "eq-creative", "name": "Creative Writing v3", "operator": "EQ-Bench", "description": "短篇创作与表达", "status": "reference_only", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "Creative Writing v3", "area": null, "officialUrl": "https://eqbench.com/creative_writing.html", "what": "官方 CSV 内嵌于数据脚本，采用 Elo；Sonnet 4.6 判胜负，风格、重复率和 rubric 不重复计票。", "usage": "写作目前不计分，只作参考。", "limits": "评分来自模型裁判；写作目前不计分。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": "2026-09-24T07:57:16.000Z", "syncedAt": "2026-10-09T00:05:24.552Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/eq-creative", "api_url": "https://aihot.news/api/site/leaderboard/sources/eq-creative", "fetched_at": "2026-10-09T04:16:53.018Z", "raw_sha256": "f89852d04fbef327f9c6fc8da03ff0e5313849434d65bf1271147de1a519e0a6" } }, "eq-longform": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "eq-longform", "name": "Longform Writing", "operator": "EQ-Bench", "description": "长篇故事的连贯性与完整性", "status": "reference_only", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "Longform Writing", "area": null, "officialUrl": "https://eqbench.com/creative_writing_longform.html", "what": "采用 overall_score_100；八章长文的情节与表达评分，裁判偏好作为局限披露。", "usage": "写作目前不计分，只作参考。", "limits": "评分来自模型裁判；写作目前不计分。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": "2026-09-24T07:57:16.000Z", "syncedAt": "2026-10-09T00:05:24.560Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/eq-longform", "api_url": "https://aihot.news/api/site/leaderboard/sources/eq-longform", "fetched_at": "2026-10-09T04:16:54.092Z", "raw_sha256": "579bff4b3c911319468958428a17a7fd0d74189907898aa7cc01ce785a6b1634" } }, "eq-emotional-v4": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "eq-emotional-v4", "name": "EQ-Bench 4 · 情绪与人际理解", "operator": "EQ-Bench", "description": "在多轮交流中理解人物情绪、隐含需求与不同的沟通偏好。", "status": "observing", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "EQ-Bench 4 · 情绪与人际理解", "area": null, "officialUrl": "https://eqbench.com/", "what": "120 个模拟人物、16 轮角色扮演，三家模型裁判轮换盲评；只采用 Elo，行为风格 traits 不作越高越好的能力分。旧 v3 不重复计票。", "usage": "观察中：公开数据和许可可用，等待新一轮评测；不延长有效期来凑分类来源数。", "limits": "英文模拟交流且尚未用人类专家验证，不能等同真人满意度或中文情商；当前成绩已超 45 天窗口。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/eq-emotional-v4", "api_url": "https://aihot.news/api/site/leaderboard/sources/eq-emotional-v4", "fetched_at": "2026-10-09T04:16:54.773Z", "raw_sha256": "de84208ec3713d81821a720531cccebcb9da24a40c475c0ac0a4ca3c8a25eec2" } }, "lechmazur-short-story": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "lechmazur-short-story", "name": "Short-Story · 短篇小说", "operator": "Lech Mazur", "description": "把固定人物、情节与风格要求自然写进一篇完整短篇小说。", "status": "observing", "brand": { "src": null, "monogram": "L", "raster": false }, "fullName": "Short-Story · 短篇小说", "area": null, "officialUrl": "https://github.com/lechmazur/writing", "what": "600–800 词短篇、10 项创作要求；同提示、三家裁判、双向比较。采用 Thurstone 主分，BT 仅作诊断，不重复计票；多档位按固定规则选取。", "usage": "观察中：补齐正式数据复用边界并解决 README 与机器成绩不同步后再接入。", "limits": "模型裁判仍有文学偏好；部分型号未完成全部故事。短篇能力不等于长篇小说或中文文笔。", "license": "未找到明确的榜单数据复用许可；GitHub 公开不等于开放授权。", "attribution": "成绩由 Lech Mazur 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/lechmazur-short-story", "api_url": "https://aihot.news/api/site/leaderboard/sources/lechmazur-short-story", "fetched_at": "2026-10-09T04:16:54.156Z", "raw_sha256": "e86a2e70446c68376d21fffc229ea916ec4717310a5b5da359a3767885c4bb49" } }, "tonebench-scripts": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "tonebench-scripts", "name": "ToneBench · 文风与脚本", "operator": "Towards AI", "description": "按指定作者文风写视频脚本，考查开场、结构、表达与口播节奏。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "ToneBench · 文风与脚本", "area": null, "officialUrl": "https://benchmark.towardsai.com/", "what": "10 个真实脚本任务、多次生成、三家裁判；只比较相同任务、rubric、裁判和运行指纹。混合模型回退行须排除，不按每列最高成绩选配置。", "usage": "观察中：数据协议和当期覆盖较好，待确认成绩再展示边界并核验纯模型配置。", "limits": "偏向 Towards AI 的单一频道文风，精确题目和参考脚本未公开；现有第一名含 Fable 5 + Opus 4.8 回退，不能作为单型号成绩。", "license": "公开 JSON 可读取，尚未找到明确的数据聚合与再展示条款。", "attribution": "成绩由 Towards AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/tonebench-scripts", "api_url": "https://aihot.news/api/site/leaderboard/sources/tonebench-scripts", "fetched_at": "2026-10-09T04:16:53.576Z", "raw_sha256": "49a1b5768ca0155d4d6c5480e17062bf7e8bb71552747dcc231db15c4c6d389b" } }, "tubelab-scriptwriting": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "tubelab-scriptwriting", "name": "TubeLab · 视频脚本", "operator": "TubeLab", "description": "面向不同视频频道写脚本，比较故事组织、文风、节奏与清晰度。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "TubeLab · 视频脚本", "area": null, "officialUrl": "https://tubelab.net/benchmark/scriptwriting", "what": "统一 TubeLab Scriptwriter、12 个任务、6 项写作质量；三家模型裁判不评自己家，失败任务保留惩罚。成本和速度不进入能力。", "usage": "观察中：等待稳定原始数据、准确评测日期和复用边界，不用展示页面代替可核验成绩。", "limits": "模型受 Scriptwriter 流程影响；OpenAI 使用最低推理档位，不能外推到其它档位。视频口播不等于电影剧本。", "license": "未确认稳定成绩导出和公开聚合许可。", "attribution": "成绩由 TubeLab 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/tubelab-scriptwriting", "api_url": "https://aihot.news/api/site/leaderboard/sources/tubelab-scriptwriting", "fetched_at": "2026-10-09T04:16:55.872Z", "raw_sha256": "9f9aa506bd53da51844e19f7d7ceb5c4c6bd20fd09062ca19d5b314d60c55a16" } }, "ncbench-author-workflow": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "ncbench-author-workflow", "name": "NC Bench · 作家工作流", "operator": "Novelcrafter", "description": "帮助作家改写、续写、总结、翻译和整理创作想法。", "status": "observing", "brand": { "src": null, "monogram": "N", "raster": false }, "fullName": "NC Bench · 作家工作流", "area": null, "officialUrl": "https://www.nc-bench.com/", "what": "只考虑独立写作和文本编辑任务的纯质量成绩；Rank Score 含价格、速度与稳定性，明确不采用。", "usage": "观察中：待核验可比的任务级原始质量成绩、真实更新日期与再展示边界。", "limits": "综合 Rank Score 混合非能力因素；不同任务覆盖不齐时不能直接用首页聚合名次。", "license": "尚未确认机器成绩与数据复用条款。", "attribution": "成绩由 Novelcrafter 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/ncbench-author-workflow", "api_url": "https://aihot.news/api/site/leaderboard/sources/ncbench-author-workflow", "fetched_at": "2026-10-09T04:16:55.355Z", "raw_sha256": "db1b940e5b463a365f51109ae1ffa59878fc5b97630ffffdf1e9667ec9a0b318" } }, "design-arena-design": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "design-arena-design", "name": "Design Arena · 视觉设计", "operator": "Design Arena / Intelligence", "description": "真人盲选网页、界面、图表和幻灯片的设计作品。", "status": "observing", "brand": { "src": null, "monogram": "D", "raster": false }, "fullName": "Design Arena · 视觉设计", "area": null, "officialUrl": "https://www.designarena.ai/leaderboard/website", "what": "仅比较 Models Arena 的基础模型；各设计任务共享家族预算，不把子标签当多个独立来源，也不与 codecategories 总分重复计票。", "usage": "观察中：正式 API 协议和展示条款已明确，待获批访问及逐项原始成绩核验。尚未采集、尚未计分。", "limits": "公开网页不替代 API 授权；模型、Agent 和 Builder 不能混排。", "license": "正式 API 数据允许个人及商业展示，须署名和可见官网链接；API 需申请获批。", "attribution": "成绩由 Design Arena / Intelligence 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/design-arena-design", "api_url": "https://aihot.news/api/site/leaderboard/sources/design-arena-design", "fetched_at": "2026-10-09T04:16:55.997Z", "raw_sha256": "18ff292465b1cfc453d8b1f33772069209fd8ca99f6c7aaca5204bd86eb29344" } }, "openvibeeval-design": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "openvibeeval-design", "name": "OpenVibeEval · 网页作品", "operator": "OpenVibeEval", "description": "对照网页作品、可访问性和盲选偏好，观察不同工具下的设计表现。", "status": "observing", "brand": { "src": null, "monogram": "O", "raster": false }, "fullName": "OpenVibeEval · 网页作品", "area": null, "officialUrl": "https://openvibeeval.com/", "what": "模型与 harness 分开保存；可访问性分不是审美分，需要核验独立盲选成绩与同运行条件才能采用。", "usage": "观察中：需补齐同条件、足量盲测的原始偏好数据及发布时间。", "limits": "模型混用不同运行工具；主表是可访问性与任务成果，不是统一审美投票。", "license": "未确认稳定成绩导出与公开聚合许可。", "attribution": "成绩由 OpenVibeEval 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/openvibeeval-design", "api_url": "https://aihot.news/api/site/leaderboard/sources/openvibeeval-design", "fetched_at": "2026-10-09T04:16:55.942Z", "raw_sha256": "a8916c1554d461bf54e74bde83610bc590160802959d4aee030fcae98497f1c9" } }, "svgbench-arena": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "svgbench-arena", "name": "SVGBench · 矢量图盲选", "operator": "SVGBench", "description": "按相同提示盲选 SVG，考查视觉清晰度与图形表现。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SVGBench · 矢量图盲选", "area": null, "officialUrl": "https://svgbench.ai/leaderboard", "what": "需要完整、已达到提示覆盖门槛的盲测分数；注册模型数量不等于有效上榜型号数。与其它 SVG 评测共用相关任务家族。", "usage": "观察中：等待稳定原始成绩、可验证更新时间与再展示边界。", "limits": "网页统计不能证明每个新型号都有足量评测；没有原始表和时间时不声称已覆盖。", "license": "未找到明确的成绩复用许可。", "attribution": "成绩由 SVGBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/svgbench-arena", "api_url": "https://aihot.news/api/site/leaderboard/sources/svgbench-arena", "fetched_at": "2026-10-09T04:16:56.444Z", "raw_sha256": "2759a3797ca40680fbf2247509654cc5863d18b984b3e989f0f48bcd288d2307" } }, "rapidata-svg": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "rapidata-svg", "name": "Rapidata SVG", "operator": "Rapidata", "description": "真人比较模型生成的矢量图，直接评价视觉偏好。", "status": "observing", "brand": { "src": "/leaderboard-sources/rapidata.svg", "monogram": "R", "raster": false }, "fullName": "Rapidata SVG · Human Preference", "area": null, "officialUrl": "https://www.benchmark.ai/", "what": "只采用同一静态 SVG 任务的 Preference Elo；Alignment、Coherence 和 Overall 不重复计票。以官方数据集固定版本的公开成绩表为准，可追溯原始人类对战。", "usage": "观察中：官方数据与许可可用，但三条 DeepSeek 滚动别名缺少固定版本证据，整张成绩暂不采纳。等待评测方补充运行版本；没有削减现役来源预算。", "limits": "SVG 绘图不能代表全部界面设计。公开原始数据不含模型版本、生成日期或 API fingerprint，无法判断滚动接口是否换版；不能用榜单更新日猜模型版本。", "license": "CC BY 4.0 · Rapidata/svg-benchmark；保留官方署名与数据链接。", "attribution": "成绩由 Rapidata 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "writing", "name": "写作与设计" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/rapidata-svg", "api_url": "https://aihot.news/api/site/leaderboard/sources/rapidata-svg", "fetched_at": "2026-10-09T04:16:56.552Z", "raw_sha256": "c77497181c7016fa278754c4d3e75fecca068af1bdf3028adfef330514ac203f" } }, "arena-vision": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "arena-vision", "name": "Arena Vision", "operator": "LMArena", "description": "真人盲选带图片的回答，观察视觉理解的实际体验。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Vision Style-Controlled", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "只读取官方数据集 vision_style_control 的 overall；不含图像生成或图像编辑。", "usage": "视觉目前不计分，只作参考。", "limits": "视觉目前不计分，只作参考。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "vision", "name": "视觉理解" } }, "upstreamAt": "2026-10-02T00:00:00.000Z", "syncedAt": "2026-10-09T00:05:21.253Z", "collected": true, "systemRows": false, "rowsNote": "按固定规则，每个公开模型采用一套代表配置；没有可采用配置的模型标为未计入并写明原因。匿名测试型号不展示，保留原榜名次，每项最多 30 个。", "provenance": { "source_url": "https://aihot.news/leaderboard/sources/arena-vision", "api_url": "https://aihot.news/api/site/leaderboard/sources/arena-vision", "fetched_at": "2026-10-09T04:16:57.158Z", "raw_sha256": "e2c04431cc65122a314ad397c2ac04a8ffcfa6adfefa1125db0d0da608245018" } }, "roboflow-vision": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "roboflow-vision", "name": "Roboflow Vision Evals", "operator": "Roboflow", "description": "读图、计数、OCR 与图像信息提取", "status": "observing", "brand": { "src": null, "monogram": "R", "raster": false }, "fullName": "Roboflow Vision Evals", "area": null, "officialUrl": "https://playground.roboflow.com/evals", "what": "六项任务共同预算；区分严格准确率与模型裁判分，不混用三次运行半极差与 95% 区间。", "usage": "观察中：新榜数据接口返回限流；旧开源仓库不含这一轮新型号，不用旧数据冒充新榜。", "limits": "观察中：新榜数据接口返回限流；旧开源仓库不含这一轮新型号，不用旧数据冒充新榜。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Roboflow 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "vision", "name": "视觉理解" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/roboflow-vision", "api_url": "https://aihot.news/api/site/leaderboard/sources/roboflow-vision", "fetched_at": "2026-10-09T04:16:57.218Z", "raw_sha256": "e85bc4e628aabd3af21fb85e7bc9a020ad8a4986c526cc74d54e667aa5ac4e0b" } }, "artificial-analysis-multilingual": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "artificial-analysis-multilingual", "name": "AA 中文／多语言", "operator": "Artificial Analysis", "description": "单独看中文和多语言表现，不再把英文综合能力重复算一次。", "status": "observing", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Artificial Analysis Multilingual Index", "area": null, "officialUrl": "https://artificialanalysis.ai/models/multilingual", "what": "只取官方多语言指数字段，不拆单语子分再投一次。", "usage": "中文与多语言目前不计分。", "limits": "中文与多语言目前不计分。", "license": "API 数据展示或分享须显著署名；再分发权与定制条款须另行取得并遵守 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "multilingual", "name": "中文与多语言" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/artificial-analysis-multilingual", "api_url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-multilingual", "fetched_at": "2026-10-09T04:16:58.346Z", "raw_sha256": "eb1f692624a6479e3efd261a6373a5e3ab6657294256df6c46d09c52a18a7b33" } }, "rws-mgate": { "run": { "id": "yiu8g6ysf3xdoq8xed0giima0", "methodologyVersion": "2026.10-capability-scale-v19", "generatedAt": "2026-10-09T00:05:49.189Z", "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "source": { "key": "rws-mgate", "name": "RWS M-GATE", "operator": "RWS", "description": "含中文的 30 种语言，考查语法与翻译", "status": "observing", "brand": { "src": null, "monogram": "R", "raster": false }, "fullName": "RWS M-GATE", "area": null, "officialUrl": "https://www.rws.com/artificial-intelligence/train-ai-data-services/research/m-gate/leaderboard/", "what": "语法 MCC 与翻译分别保存；分词效率不进入能力分数。", "usage": "观察中：浏览器可见新模型，但机器接口返回空内容；缺少可验证的成绩发布时间与稳定导出。", "limits": "观察中：浏览器可见新模型，但机器接口返回空内容；缺少可验证的成绩发布时间与稳定导出。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 RWS 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": { "key": "multilingual", "name": "中文与多语言" } }, "upstreamAt": null, "syncedAt": null, "collected": false, "systemRows": false, "rowsNote": null, "provenance": { "source_url": "https://aihot.news/leaderboard/sources/rws-mgate", "api_url": "https://aihot.news/api/site/leaderboard/sources/rws-mgate", "fetched_at": "2026-10-09T04:16:58.346Z", "raw_sha256": "f003bca898c35281969041d6b60f7907e6d27b9e5cca8d657640ae44883e5221" } } }, "methodology": { "run_id": "yiu8g6ysf3xdoq8xed0giima0", "commit": "3343fe2b20db4be7269113752d82d3992fc52b6b", "board_groups": { "coding": "coding", "reasoning": "reasoning", "professional": "professional", "knowledge": "knowledge" }, "budgets": [ { "key": "coding", "name": "编程", "weight": 0.4, "sources": [ "Terminal-Bench 4.0 · Artificial Analysis", "Terminal-Bench 4.0 · Vals AI", "LiveBench Coding · LiveBench", "SciCode · Artificial Analysis", "Code Migration · Vals AI", "ProgramBench · Vals AI", "Vibe Code Bench v1.1 · Vals AI", "Vibe Code Bench 1–100 · Vals AI", "CyberBench Patch v1.1 · Vals AI" ] }, { "key": "reasoning", "name": "科研", "weight": 0.2, "sources": [ "Humanity's Last Exam · Artificial Analysis", "FrontierMath T1–3 · Epoch AI", "LiveBench Reasoning · LiveBench", "CritPt · Artificial Analysis", "Chess Puzzles · Epoch AI", "Mystery Game Puzzles · Epoch AI", "MysteryMechanism · Vals AI", "Terminal-Bench Science · Vals AI", "Terminal-Bench Science · Artificial Analysis" ] }, { "key": "professional", "name": "专业办公", "weight": 0.2, "sources": [ "GDPval-AA · Artificial Analysis", "AA-Briefcase · Artificial Analysis", "AutomationBench-AA · Artificial Analysis", "Finance Agent · Vals AI", "APEX-Agents · Mercor", "Harvey 法律 Agent 基准 · Vals AI", "τ³-Banking · Mercor", "LiveBench Data Analysis · LiveBench", "MedCode · Vals AI", "AA-Analyst Agent · Artificial Analysis" ] }, { "key": "knowledge", "name": "知识问答", "weight": 0.15, "sources": [ "AA-Omniscience · Artificial Analysis", "SimpleQA Verified · Epoch AI", "AA-LCR · Artificial Analysis", "MLCR-AA · Artificial Analysis", "LiveBench Language · LiveBench", "LiveBench IF · LiveBench", "BullshitBench v2 · BullshitBench" ] }, { "key": "preference", "name": "真人盲选", "weight": 0.05, "sources": [ "Arena Text · LMArena" ] } ], "sources": [ { "key": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0 · AA", "operator": "Artificial Analysis", "description": "在命令行环境里独立完成一整件工程任务。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Terminal-Bench 4.0（Artificial Analysis 复跑）", "area": "编程", "officialUrl": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "what": "取 Artificial Analysis 智能指数中 Terminal-Bench 4.0 的通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域，与 Vals 复跑的 Terminal-Bench 4.0 同属一类，两家合起来算一票。", "limits": "终端任务对运行环境和时限设置敏感；厂商接口回退的运行照常计入并标注（AA 不公布受影响比例）。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "weight": null, "status_label": "参与排名" }, { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals", "operator": "Vals AI", "description": "Vals 用同一套 Agent 流程给所有模型复跑 Terminal-Bench 4.0。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Terminal-Bench 4.0（Vals AI 复跑）", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-4", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域，与 AA 复跑的 Terminal-Bench 4.0 同属一类，两家合起来算一票。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/terminal-bench-4", "weight": null, "status_label": "参与排名" }, { "key": "livebench-coding", "name": "LiveBench · 编程综合", "operator": "LiveBench", "description": "Coding + Agentic Coding 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 编程综合", "area": "编程", "officialUrl": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "what": "读取 Coding 和 Agentic Coding 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：只计 Coding 一项（编程领域的写代码一类）。Agentic Coding 对部分厂商使用原生工具调用提示，流程不统一，暂不计分。", "limits": "两项均分衡量组合表现，不能当成 Coding 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": "coding", "category": "coding", "url": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-scicode", "name": "SciCode · AA", "operator": "Artificial Analysis", "description": "把物理、化学、生物等科研问题写成能运行的计算代码。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "SciCode（Artificial Analysis 复跑）", "area": "编程", "officialUrl": "https://artificialanalysis.ai/evaluations/scicode", "what": "取 Artificial Analysis 智能指数中 SciCode 的通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的科学计算一类。", "limits": "题目来自公开数据集，可能存在训练污染；题集正在接受公开复核。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://artificialanalysis.ai/evaluations/scicode", "weight": null, "status_label": "参与排名" }, { "key": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "description": "把整套程序从一种语言或框架迁移到另一种。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Code Migration", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/code-migration", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的程序迁移一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/code-migration", "weight": null, "status_label": "参与排名" }, { "key": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "description": "只给可执行文件，从零重写出行为相同的程序。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "ProgramBench", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/programbench", "what": "取 Vals 专项页 Overall 通过率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：编程领域的程序重建一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/programbench", "weight": null, "status_label": "参与排名" }, { "key": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "description": "从需求搭建网页应用，检查实际功能。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vibe Code Bench v1.1", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/vibe-code", "what": "50 个应用开发场景，统一 OpenHands，浏览器评测按验收步骤检查功能。", "usage": "计分：应用开发；与 Vibe Code Bench 1–100 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 OpenHands 成绩；厂商专用 Agent 不计。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/vibe-code", "weight": null, "status_label": "参与排名" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "description": "在已有应用上连续开发，同时保持旧功能。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vibe Code Bench 1–100", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/vcb-1-100", "what": "50 个连续开发测试场景；统一 OpenHands、每场景 10 小时预算，浏览器验证新增功能与回归。", "usage": "计分：应用开发；与 Vibe Code Bench 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "版本 1.0；连续开发总体成绩。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/vcb-1-100", "weight": null, "status_label": "参与排名" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "description": "修补真实开源项目的安全漏洞。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "CyberBench Patch v1.1", "area": "编程", "officialUrl": "https://www.vals.ai/benchmarks/cyber", "what": "56 个 Patch 任务，统一 mini-swe-agent；编译、原漏洞复测及隐藏输入与维护者修复对照。", "usage": "计分：漏洞修补一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 Patch，不混入 PoC 或总分；回退比例按 Patch 的 56 题计算。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.vals.ai/benchmarks/cyber", "weight": null, "status_label": "参与排名" }, { "key": "deepswe-v1-1", "name": "DeepSWE v1.1", "operator": "DataCurve", "description": "在统一的编程助手环境里改仓库、修缺陷，比较的是写代码的模型本身。", "status": "observing", "brand": { "src": "/leaderboard-sources/deepswe.svg", "monogram": "D", "raster": false }, "fullName": "DeepSWE v1.1", "area": "编程", "officialUrl": "https://deepswe.datacurve.ai/", "what": "只取 pass@1；统一 harness 内按预先固定的推理档位优先级选代表配置。", "usage": "观察中：9 月 3 日后没有再加入新模型，且 Epoch 指出部分题目判分有缺陷；恢复更新并修好后再计分。", "limits": "仍是模型＋mini-swe-agent 的系统表现；大版本变化必须隔离快照。", "license": "官方公开结构化结果；仅管理员私有观察，公开前复核许可", "attribution": "成绩由 DataCurve 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://deepswe.datacurve.ai/", "weight": null, "status_label": "观察中" }, { "key": "epoch-mirrorcode", "name": "MirrorCode", "operator": "Epoch AI", "description": "长时间的软件复现", "status": "observing", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "MirrorCode", "area": "编程", "officialUrl": "https://epoch.ai/benchmarks/mirrorcode", "what": "已发布数据仅少量型号，且运行时间很长；样本和运行条件不足以形成分类共识。", "usage": "观察中：只测了少数模型，测满 20 个模型再计分。", "limits": "只测了少数模型，刻度定不准。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://epoch.ai/benchmarks/mirrorcode", "weight": null, "status_label": "观察中" }, { "key": "taptap-maker", "name": "TapTap Maker", "operator": "TapTap Maker", "description": "在真实游戏引擎里用 Lua 写出能跑起来的功能。分数由引擎执行判定，不用另一个模型打分。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/taptap-maker.svg", "monogram": "T", "raster": false }, "fullName": "TapTap Maker Benchmark", "area": "编程", "officialUrl": "https://maker.taptap.cn/leaderboard/", "what": "只取官方 main_board 的 L2 通过率；成本、速度、Held-out 和分类分不再次投票。同一基础模型按固定推理档位优先级选代表行，选择不看分数。", "usage": "仅供参考：现有结果混合不同 agent driver，不能作为相同执行流程下的底层型号比较。", "limits": "单引擎、单语言，不能推广到其他技术栈；不同模型可能使用不同 agent 驱动；闭源 harness 不能逐题复算。 当前采用的执行驱动不同，不计入单型号排序。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以易玩／TapTap 条款为准", "attribution": "成绩由 TapTap Maker 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://maker.taptap.cn/leaderboard/", "weight": null, "status_label": "仅供参考" }, { "key": "terminal-bench-4", "name": "Terminal-Bench 4", "operator": "Harbor / Terminal-Bench", "description": "保留模型搭配不同 Agent 在终端任务中的原始成绩，供交叉核对。", "status": "reference_only", "brand": { "src": null, "monogram": "H", "raster": false }, "fullName": "Terminal-Bench 4 · 系统参考", "area": "编程", "officialUrl": "https://www.tbench.ai/", "what": "固定 Terminal-Bench 4.0，分别保留模型、Agent 版本、推理档位和运行次数；不把不同工具系统解释为统一模型条件。", "usage": "原始成绩参考：锁定官方提交清单与数据版本后自动采集，逐条保留不同系统配置，不选模型代表分。只供交叉核对，不进入综合或编程排名。", "limits": "不同模型使用 Claude Code、Codex、GrokBuild 或 mini-SWE 等不同工具，只供参考；AA 和 Vals 各自统一复跑的 Terminal-Bench 4.0 已计分。", "license": "Apache 2.0 · 官方 harbor-framework/terminal-bench 仓库内的成绩提交；保留署名、协议与修改说明。", "attribution": "成绩由 Harbor / Terminal-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.tbench.ai/", "weight": null, "status_label": "仅供参考" }, { "key": "hyper-tau", "name": "Hyper-τ-bench", "operator": "Sierra", "description": "根据业务资料构建并测试能工作的客户服务代理。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "Hyper-τ-bench", "area": "编程", "officialUrl": "https://sierra.ai/blog/hyper-t-bench-evaluating-agents-that-build-agents", "what": "衡量代理开发与调试，和直接执行客户服务的 τ³-Banking 分开；固定开发工具、运行时长与验收环境，不能跨系统挑最高成绩。", "usage": "观察中：可补编程与业务系统开发，先完成原始成绩、许可和固定运行条件审查；不因新发布就自动计分。", "limits": "观察中：可补编程与业务系统开发，先完成原始成绩、许可和固定运行条件审查；不因新发布就自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Sierra 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://sierra.ai/blog/hyper-t-bench-evaluating-agents-that-build-agents", "weight": null, "status_label": "观察中" }, { "key": "swe-rebench", "name": "SWE-rebench", "operator": "Nebius", "description": "持续收集真实仓库问题，比较模型在多种编程语言中的软件修复能力。", "status": "observing", "brand": { "src": null, "monogram": "N", "raster": false }, "fullName": "SWE-rebench", "area": "编程", "officialUrl": "https://swe-rebench.com/", "what": "固定任务窗口、ReAct 环境和 text/tools 交互协议，采用多次运行的平均解决率与标准误；不把完整 Agent 产品与基础模型混为同一对象。", "usage": "候选观察：真实仓库任务值得补充，等待可验证运行日期、稳定成绩协议与再展示边界；尚未启用自动采集或计分。", "limits": "滚动题目、运行环境和任务窗口影响可比性；网站近期维护不等于所有模型重新测试。", "license": "官方 Hugging Face CC BY 4.0 数据集是任务题库；尚未确认网页最终成绩的稳定导出与再展示边界。", "attribution": "成绩由 Nebius 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://swe-rebench.com/", "weight": null, "status_label": "观察中" }, { "key": "lhtb", "name": "LHTB", "operator": "Tencent HY / Lehigh 等", "description": "在较长时间内持续使用终端，完成复杂工程与工具任务。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "Long-Horizon Terminal-Bench", "area": "编程", "officialUrl": "https://zli12321.github.io/LHTB/leaderboard.html", "what": "90 分钟、2 小时和 3 小时预算分别比较；修复前后判题协议不能混排，也不能把不同 Agent 的最好结果当统一模型成绩。", "usage": "候选观察：等待明确标注修复协议的新批次，现有旧分不自动采集到公开榜，也不计分；不据此断言每条旧结果都利用了漏洞。", "limits": "官方披露过判题信息泄漏；当前不能证明已有成绩属于隔离判题后的同一协议。部分任务与其他专业评测的来源重叠待核。", "license": "Apache 2.0 · 官方 IntelligenceLab/LHTB-leaderboard 成绩数据集；保留署名和变更说明。", "attribution": "成绩由 Tencent HY / Lehigh 等 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://zli12321.github.io/LHTB/leaderboard.html", "weight": null, "status_label": "观察中" }, { "key": "frontiercode", "name": "FrontierCode", "operator": "Cognition", "description": "真实仓库任务的质量、测试与修改范围", "status": "observing", "brand": { "src": null, "monogram": "C", "raster": false }, "fullName": "FrontierCode", "area": "编程", "officialUrl": "https://cognition.com/frontiercode", "what": "固定 v1.1 和官方运行条件；质量评分和测试成功共同构成一项任务证据。", "usage": "观察中：尚无确认稳定的成绩导出与数据使用边界。", "limits": "观察中：尚无确认稳定的成绩导出与数据使用边界。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Cognition 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://cognition.com/frontiercode", "weight": null, "status_label": "观察中" }, { "key": "frontierswe-v2", "name": "FrontierSWE v2", "operator": "Proximal Labs", "description": "长时间运行的工程实现与研究任务", "status": "observing", "brand": { "src": null, "monogram": "P", "raster": false }, "fullName": "FrontierSWE v2", "area": "编程", "officialUrl": "https://www.frontierswe.com/", "what": "只比较 v2 的统一 Proximus harness，不与旧版和 best-of 不同条件混合。", "usage": "观察中：长任务新型号覆盖仍较少，稳定原始数据入口和运行配置需要补齐。", "limits": "观察中：长任务新型号覆盖仍较少，稳定原始数据入口和运行配置需要补齐。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Proximal Labs 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://www.frontierswe.com/", "weight": null, "status_label": "观察中" }, { "key": "weirdml", "name": "WeirdML", "operator": "WeirdML", "description": "在陌生数据上实现机器学习方案", "status": "observing", "brand": { "src": null, "monogram": "W", "raster": false }, "fullName": "WeirdML", "area": "编程", "officialUrl": "https://htihle.github.io/weirdml.html", "what": "按实际任务清单和固定 GPU 时间比较，保留五轮迭代的计分协议。", "usage": "观察中：官方 CSV 可读，但成绩许可和逐型号运行时间尚未完整核实。", "limits": "观察中：官方 CSV 可读，但成绩许可和逐型号运行时间尚未完整核实。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 WeirdML 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "coding", "category": "coding", "url": "https://htihle.github.io/weirdml.html", "weight": null, "status_label": "观察中" }, { "key": "artificial-analysis-hle", "name": "HLE · AA", "operator": "Artificial Analysis", "description": "横跨学科的专家级难题，不用工具作答。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Humanity’s Last Exam（Artificial Analysis 复跑）", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "what": "取 Artificial Analysis 智能指数中 Humanity’s Last Exam 的正确率（不用工具）。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的专家难题一类。", "limits": "部分题目答案存在争议；正确率普遍偏低，接近 0 的成绩只说明“至多这么强”。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "weight": null, "status_label": "参与排名" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3", "operator": "Epoch AI", "description": "研究级数学推理", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "FrontierMath v2 · Tiers 1–3", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "what": "只用 v2 的 Tiers 1–3；旧版题目和 Tier 4 不混入同一指标。", "usage": "计分：科研领域的研究级数学一类。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "weight": null, "status_label": "参与排名" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学", "operator": "LiveBench", "description": "Reasoning + Mathematics 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 推理与数学", "area": "科研", "officialUrl": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "what": "读取 Reasoning 和 Mathematics 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：只计 Reasoning 一项（科研领域的逻辑一类）。Mathematics 头部已接近满分，暂不计分。", "limits": "两项均分衡量组合表现，不能当成 Reasoning 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": "reasoning", "category": "reasoning", "url": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-critpt", "name": "CritPt · AA", "operator": "Artificial Analysis", "description": "研究级物理问题，考验专业推导。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "CritPt（Artificial Analysis 复跑）", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/critpt", "what": "取 Artificial Analysis 智能指数中 CritPt 的正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的研究级物理一类。", "limits": "题量较少，正确率普遍很低；题集正在接受公开复核。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://artificialanalysis.ai/evaluations/critpt", "weight": null, "status_label": "参与排名" }, { "key": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "description": "根据文字棋局寻找正确走法", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "Chess Puzzles", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/chess-puzzles", "what": "生成的棋局以 Stockfish 标准答案判分；输入为棋局文本，不归入视觉理解。", "usage": "计分：科研领域的棋类与游戏一类（与 Mystery Game Puzzles 合起来算一票）。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://epoch.ai/benchmarks/chess-puzzles", "weight": null, "status_label": "参与排名" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "description": "从陌生规则推导正确行动", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "Mystery Game Puzzles", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/mystery-game-puzzles", "what": "只取 Epoch 标准协议的 mean_score；题目不公开这一限制明确保留。", "usage": "计分：科研领域的棋类与游戏一类（与 Chess Puzzles 合起来算一票）。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://epoch.ai/benchmarks/mystery-game-puzzles", "weight": null, "status_label": "参与排名" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "description": "自己设计实验，去推断一套隐藏的规律。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MysteryMechanism（Vals AI）", "area": "科研", "officialUrl": "https://www.vals.ai/benchmarks/mysterymechanism", "what": "取 Vals 专项页 Overall 成绩。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：科研领域的实验推断一类。", "limits": "参评模型较少；Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://www.vals.ai/benchmarks/mysterymechanism", "weight": null, "status_label": "参与排名" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals", "operator": "Vals AI", "description": "在终端完成科学工作流与分析。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Terminal-Bench Science · Vals", "area": "科研", "officialUrl": "https://www.vals.ai/benchmarks/terminal-bench-science", "what": "70 个专家审核的科学终端任务，统一 Mini-SWE-agent，由任务验证器检查结果。", "usage": "计分：科学工作流；与 AA 复跑的同一题库共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取版本 0.1 的 Mini-SWE-agent 成绩；Terminus 2 不混入。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://www.vals.ai/benchmarks/terminal-bench-science", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA", "operator": "Artificial Analysis", "description": "在终端完成科学工作流与分析。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Terminal-Bench Science · AA", "area": "科研", "officialUrl": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "what": "70 个 Harbour 科学终端任务，统一 mini-swe-agent，每题三次运行的平均 pass@1。", "usage": "计分：科学工作流；与 Vals 复跑的同一题库共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "取 terminalBenchScience 总体成绩，不混用领域切片。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "weight": null, "status_label": "参与排名" }, { "key": "epoch-frontiermath-tier4", "name": "FrontierMath v2 · Tier 4", "operator": "Epoch AI", "description": "更高难度的数学研究题", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "FrontierMath v2 · Tier 4", "area": "科研", "officialUrl": "https://epoch.ai/benchmarks/frontiermath-tier-4-v2", "what": "独立保存 Tier 4，和 Tiers 1–3 共用数学证据预算；保留样本误差。", "usage": "Tier 4 题量太少，只作参考，不计分。", "limits": "Tier 4 题量太少，只作参考，不计分。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://epoch.ai/benchmarks/frontiermath-tier-4-v2", "weight": null, "status_label": "仅供参考" }, { "key": "hle", "name": "Humanity’s Last Exam", "operator": "CAIS / Scale AI", "description": "横跨学科的高难度学术知识与推理。", "status": "observing", "brand": { "src": null, "monogram": "C", "raster": false }, "fullName": "Humanity’s Last Exam", "area": "知识", "officialUrl": "https://labs.scale.com/leaderboard/humanitys_last_exam", "what": "固定题集、工具条件和裁判；HLE、HLE-Rolling 与带工具结果分别比较。AA v4.3 已包含 HLE，将来若有可用专项成绩，只用于分类，不在综合榜重复计票。", "usage": "观察中：AA 统一复跑的 HLE 已计分；官方榜单接受各家自报成绩，暂不计分。", "limits": "观察中：AA 统一复跑的 HLE 已计分；官方榜单接受各家自报成绩，暂不计分。", "license": "评测代码 MIT；官方榜单成绩再展示边界尚待确认。", "attribution": "成绩由 CAIS / Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://labs.scale.com/leaderboard/humanitys_last_exam", "weight": null, "status_label": "观察中" }, { "key": "simplebench", "name": "SimpleBench", "operator": "SimpleBench", "description": "用日常常识和逻辑问题，检验模型能否识别问题中的实际约束。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SimpleBench", "area": "科研", "officialUrl": "https://simple-bench.com/", "what": "区分选择题与开放回答，固定 AVG@5、温度和提示协议；不把加入榜单日期当逐题评测时间。", "usage": "候选观察：新品覆盖及时，等待官网成绩复用边界与可冻结的数据版本说明；不套用代码许可，尚未自动采集或计分。", "limits": "日常常识题具有补充价值，但题目代表性与选择题/开放回答条件均影响解释。", "license": "样题与评测代码仓库 MIT；官网独立发布的最新成绩脚本再展示边界尚未确认。", "attribution": "成绩由 SimpleBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://simple-bench.com/", "weight": null, "status_label": "观察中" }, { "key": "matharena-lean", "name": "MathArena · ArXivLean", "operator": "MathArena / ETH Zurich", "description": "用 Lean 验证数学证明", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · ArXivLean", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://matharena.ai/", "weight": null, "status_label": "观察中" }, { "key": "matharena-broken", "name": "MathArena · BrokenArXiv", "operator": "MathArena / ETH Zurich", "description": "发现和修复数学证明中的错误", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · BrokenArXiv", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://matharena.ai/", "weight": null, "status_label": "观察中" }, { "key": "matharena-math", "name": "MathArena · ArXivMath", "operator": "MathArena / ETH Zurich", "description": "从近期数学论文提炼的新题", "status": "observing", "brand": { "src": null, "monogram": "M", "raster": false }, "fullName": "MathArena · ArXivMath", "area": "科研", "officialUrl": "https://matharena.ai/", "what": "按单独试卷和版本读取原始成绩，不使用跨题库拟合的 Expected performance 综合分。", "usage": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "limits": "观察中：榜单成绩再展示边界尚未明确，暂不采集进正式分。", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "attribution": "成绩由 MathArena / ETH Zurich 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://matharena.ai/", "weight": null, "status_label": "观察中" }, { "key": "arc-agi-2", "name": "ARC-AGI-2", "operator": "ARC Prize", "description": "从陌生规则中学习和泛化", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "ARC-AGI-2", "area": "科研", "officialUrl": "https://arcprize.org/leaderboard", "what": "固定标准 harness，与提供商专门优化的 adapter 成绩分开。", "usage": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "limits": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 ARC Prize 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://arcprize.org/leaderboard", "weight": null, "status_label": "观察中" }, { "key": "arc-agi-3", "name": "ARC-AGI-3", "operator": "ARC Prize", "description": "从陌生规则中学习和泛化", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "ARC-AGI-3", "area": "科研", "officialUrl": "https://arcprize.org/leaderboard", "what": "固定标准 harness，与提供商专门优化的 adapter 成绩分开。", "usage": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "limits": "观察中：官网 Terms 对系统性采集和聚合要求书面许可；当前官方成绩的聚合边界尚待落实。标准 harness 与 Provider Adapter 分开，不拿专用适配器最高分代表普通模型。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 ARC Prize 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "reasoning", "category": "reasoning", "url": "https://arcprize.org/leaderboard", "weight": null, "status_label": "观察中" }, { "key": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "description": "按真实职业交付件出题，两两比较产出质量。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "GDPval-AA（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/gdpval-aa", "what": "取 Artificial Analysis 智能指数中 GDPval-AA 的 Elo 等级分。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域，与 AA-Briefcase 同属职业交付一类，两项合起来算一票。", "limits": "等级分来自模型裁判两两比较，不是正确率；它的刻度按同一批模型在其他领域的水平对齐，不能自定尺度。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://artificialanalysis.ai/evaluations/gdpval-aa", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "description": "办公知识工作的长任务，两两比较交付结果。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Briefcase（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/aa-briefcase", "what": "取 Artificial Analysis 智能指数中 AA-Briefcase 的 Elo 等级分。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域，与 GDPval-AA 同属职业交付一类，两项合起来算一票。", "limits": "等级分来自模型裁判两两比较，不是正确率；它的刻度按同一批模型在其他领域的水平对齐，不能自定尺度。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://artificialanalysis.ai/evaluations/aa-briefcase", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-automationbench", "name": "AutomationBench · AA", "operator": "Artificial Analysis", "description": "跨办公软件完成一整条业务自动化流程。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AutomationBench-AA（Artificial Analysis）", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/automationbench-aa", "what": "取 Artificial Analysis 智能指数中 AutomationBench-AA 的完成率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的业务流程一类。", "limits": "在 AA 的统一环境中运行，与 Zapier 官方榜单的私有测试集不同。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://artificialanalysis.ai/evaluations/automationbench-aa", "weight": null, "status_label": "参与排名" }, { "key": "vals-finance-agent", "name": "Vals Finance Agent", "operator": "Vals AI", "description": "金融分析师的日常办公题，看研报、建模这类工作做得怎样。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Vals Finance Agent", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/fabv2", "what": "取 Vals 专项页 Overall 正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的金融一类。", "limits": "Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/fabv2", "weight": null, "status_label": "参与排名" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1", "operator": "Mercor", "description": "投行、咨询和法律里的长任务，看模型当助手能不能把一件完整的工作做完。", "status": "ranked", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "fullName": "Mercor APEX-Agents 1.1", "area": "专业办公", "officialUrl": "https://www.mercor.com/apex/apex-agents-leaderboard/", "what": "固定 1.1 数据集和官方 Loop 环境，采用 Pass@1 及其对应误差，按预先固定的推理档位选配置。Mean Score 是评分项平均完成度，不混作任务完成率。", "usage": "计分：专业办公领域的投行、咨询与律所长任务一类。", "limits": "限投行、咨询、法律任务；原页未提供逐型号评测日期，少量运行缺失的原因尚未公开。不能将此分数解释为所有办公工作的可靠性。", "license": "官方公开结果；数据与代码许可不等于榜单再分发授权", "attribution": "成绩由 Mercor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.mercor.com/apex/apex-agents-leaderboard/", "weight": null, "status_label": "参与排名" }, { "key": "vals-harvey-lab", "name": "Harvey Legal Agent Benchmark", "operator": "Vals AI / Harvey", "description": "处理法律资料并交付 Word、Excel、幻灯片等可审阅文件。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Harvey Legal Agent Benchmark", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/hlab", "what": "固定 Harvey 私有测试集、文件工具和 docx、xlsx、pptx 技能，禁用互联网；取两名裁判的整任务通过率均值。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的法律一类。", "limits": "整项任务通过率普遍很低；Vals 公布每次运行有多少题被厂商接口转给了同厂商的旧型号：有回退的成绩照常计入并在模型页标注，受影响题目过半的不计。同属 Vals 的几项评测不能当作多家独立机构。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI / Harvey 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/hlab", "weight": null, "status_label": "参与排名" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor", "operator": "Mercor", "description": "在统一工具环境里办理银行客服业务：查政策、找产品、改账户。", "status": "ranked", "brand": { "src": "/leaderboard-sources/mercor.svg", "monogram": "M", "raster": false }, "fullName": "τ³-Banking（Mercor 复跑）", "area": "专业办公", "officialUrl": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "what": "取 Mercor 给所有模型统一复跑的 pass@1。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：专业办公领域的客服一类。", "limits": "依赖模拟用户与工具协议，只代表银行客服场景；Sierra 官方榜单接受厂商自报成绩，不计分。", "license": "官方公开结果；题目与评测框架来自 Sierra tau2-bench（MIT）", "attribution": "成绩由 Mercor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "weight": null, "status_label": "参与排名" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析", "operator": "LiveBench", "description": "表格重排、列类型标注、表连接预测等数据处理任务。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 数据分析", "area": "专业办公", "officialUrl": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "what": "读取 Data Analysis 分类的未舍入成绩。", "usage": "计分：专业办公领域的表格数据处理一类。", "limits": "题目每期更新，格式要求严格；组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "成绩由 LiveBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "weight": null, "status_label": "参与排名" }, { "key": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "description": "根据住院记录给出医疗诊断编码。", "status": "ranked", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MedCode", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/medcode", "what": "去标识化住院记录，由两名认证编码员标注并解决分歧；统一温度 1、最多 30k 输出 token，核对 ICD-10-CM 主次诊断编码。", "usage": "计分：医疗编码一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "取总体编码得分，不等同于整份病例完全正确率；不混用主次诊断切片。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/medcode", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "description": "可靠地分析数据并回答专业问题。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Analyst Agent", "area": "专业办公", "officialUrl": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "what": "80 个私有问题、14 个领域，统一 Stirrup 工具环境；每题独立运行五次，五次全部成功才通过（pass^5）。", "usage": "计分：数据分析；与 LiveBench Data Analysis 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "pass^5 不是五次至少成功一次的 pass@5。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "weight": null, "status_label": "参与排名" }, { "key": "tau-banking", "name": "τ³-Banking", "operator": "Sierra", "description": "在统一工具环境中办理银行业务，检验规则理解、客户沟通与任务完成。", "status": "observing", "brand": { "src": "/leaderboard-sources/sierra.png", "monogram": "S", "raster": true }, "fullName": "τ³-Banking", "area": "专业办公", "officialUrl": "https://taubench.com/leaderboard/?benchmark=knowledge", "what": "只比较官方 standard 银行业务、v1.0.1、alltools 和 GPT-5.2 用户模拟器；采用 pass@1。不同工具、提示、模拟器、语音模式与自定义系统不混排。", "usage": "观察中：Mercor 统一复跑的 τ³-Banking 已计分；官方榜单接受厂商自报成绩，不计分。", "limits": "成绩属于固定银行任务系统，依赖模拟用户与工具协议；不是所有职业的通用正确率。", "license": "MIT · Sierra Research tau2-bench 官方仓库；保留来源署名。", "attribution": "成绩由 Sierra 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://taubench.com/leaderboard/?benchmark=knowledge", "weight": null, "status_label": "观察中" }, { "key": "zapier-automation", "name": "AutomationBench", "operator": "Zapier", "description": "跨办公软件完成自动化工作", "status": "observing", "brand": { "src": null, "monogram": "Z", "raster": false }, "fullName": "AutomationBench", "area": "专业办公", "officialUrl": "https://zapier.com/benchmarks", "what": "只比较同一官方私有测试集的配置；Fable 5.1 混用 Opus 5 的回退行必须排除。", "usage": "观察中：AA 统一复跑的 AutomationBench 已计分；官方私有排行榜数据暂不计分。", "limits": "观察中：AA 统一复跑的 AutomationBench 已计分；官方私有排行榜数据暂不计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Zapier 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://zapier.com/benchmarks", "weight": null, "status_label": "观察中" }, { "key": "officeqa-pro", "name": "OfficeQA Pro", "operator": "Databricks", "description": "从大量真实财务文件中检索、计算并回答问题。", "status": "observing", "brand": { "src": null, "monogram": "D", "raster": false }, "fullName": "OfficeQA Pro", "area": "专业办公", "officialUrl": "https://github.com/databricks/officeqa", "what": "Pro 固定 133 题，与 Full 246 题及 Pro V2 的 90 题分开；端到端检索与直接提供答案所在页面属于不同条件。Claude Agent SDK、Codex SDK、Antigravity CLI 的成绩属于各自运行系统，不能混作统一基础模型评测。它衡量文件理解，不代表办公文件的产出质量。", "usage": "观察中：官方继续更新，但目前主要比较不同产品运行系统，原始汇总协议仍待核实；保留资料理解参考，不将不同工具系统的成绩直接归给基础模型。", "limits": "代码及题库许可不自动覆盖所有外部成绩；新旧题集、端到端与答案页面直供条件不能混排，当前没有自动采集成绩。", "license": "代码 Apache 2.0；题目数据 CC BY-SA 4.0，现迁至需要申请访问的 Hugging Face 数据集。", "attribution": "成绩由 Databricks 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://github.com/databricks/officeqa", "weight": null, "status_label": "观察中" }, { "key": "mcp-atlas", "name": "MCP Atlas", "operator": "Scale AI", "description": "通过真实 MCP 服务检索资料、操作文档与数据库，完成跨软件任务。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "MCP Atlas", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/mcp_atlas", "what": "固定 2026 年 4 月调整后的运行协议和 100 次工具调用预算；完整 1,000 题与公开 500 题分开。只取任务通过率，不把回答条件覆盖率重复计票；裁判、重试和工具版本需随成绩批次锁定。", "usage": "观察中：可补跨办公软件的工具执行证据；待确认稳定成绩导出、实际评测批次和再展示边界，目前只收录来源，不自动采集或计分。", "limits": "网页旧方法说明和实时成绩可能不同步；公开题集代码不是实时成绩导出。与 Scale 的其他榜共属同一机构，不增加独立机构数。", "license": "官方环境代码采用 MIT；该许可不自动覆盖实时网页成绩，榜单数据再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://labs.scale.com/leaderboard/mcp_atlas", "weight": null, "status_label": "观察中" }, { "key": "financebenchmark", "name": "FinanceBenchmark", "operator": "FinanceBenchmark", "description": "完成金融定价、资本计算和风控任务，用数值与代码执行结果核验。", "status": "observing", "brand": { "src": null, "monogram": "F", "raster": false }, "fullName": "FinanceBenchmark", "area": "专业办公", "officialUrl": "https://finbenchmark.ai/", "what": "题集 v2、v3 与运行环境版本分开；保留每题三次尝试的条件。官网 Tasks passed 指至少一次成功，却同时称为 Pass@1，不能与单次平均正确率混用；Finance Index 也是另一个汇总指标。", "usage": "观察中：可补确定答案与代码验证的金融证据；指标口径仍需核对，官方 Sol 结果 JSON 下载本次返回 404，当前只收录来源，不自动采集或计分。", "limits": "三次至少一次成功不等于单次成功率；区间必须对应正确指标，滚动模型别名也需核验版本。题库和下载按钮公开不等于原始成绩已能获取。", "license": "官方仓库代码为 MIT，网站要求保留 FinanceBenchmark 引用；未确认实时成绩完整导出的使用边界，不将代码许可替代榜单数据许可。", "attribution": "成绩由 FinanceBenchmark 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://finbenchmark.ai/", "weight": null, "status_label": "观察中" }, { "key": "prbench-legal", "name": "PRBench · Legal", "operator": "Scale AI", "description": "回答真实法律工作中的复杂问题，检验专业判断和论证质量。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "PRBench · Legal", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/prbench-legal", "what": "固定 Legal 完整 500 题或 Hard 250 题之一；搜索工具、推理档位和裁判版本保持一致。采用该题集的 rubric 汇总分，不把完整集与困难子集重复计票，也不将专业问答描述为完整文件交付。", "usage": "观察中：可补法律专业判断；先核实题集、工具条件和成绩批次，目前只收录来源，不自动采集或计分。", "limits": "由模型裁判评审，存在裁判偏好；同机构的法律和金融专项应共用影响边界，不能作为两家独立证据。它不直接衡量 Word、Excel 或幻灯片的制作能力。", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://labs.scale.com/leaderboard/prbench-legal", "weight": null, "status_label": "观察中" }, { "key": "prbench-finance", "name": "PRBench · Finance", "operator": "Scale AI", "description": "回答真实金融决策问题，检验专业判断、推导和风险说明。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "PRBench · Finance", "area": "专业办公", "officialUrl": "https://labs.scale.com/leaderboard/prbench-finance", "what": "固定 Finance 完整 600 题或 Hard 300 题之一；搜索工具、推理档位和裁判版本保持一致。采用该题集的 rubric 汇总分，不将完整集和困难子集重复计票，也不把专业问答当作端到端办公任务。", "usage": "观察中：可补金融专业判断；待核实题集、工具条件、稳定导出和再展示边界，目前只收录来源，不自动采集或计分。", "limits": "由模型裁判评审，不代表已经完成财务建模或表格制作；与 PRBench Legal、MCP Atlas 同属 Scale，不能靠增加榜单数量提高机构权重。", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "attribution": "成绩由 Scale AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://labs.scale.com/leaderboard/prbench-finance", "weight": null, "status_label": "观察中" }, { "key": "spreadsheetbench-2", "name": "SpreadsheetBench 2", "operator": "SpreadsheetBench / Renmin University / AfterQuery", "description": "完成财务建模、修复工作簿和数据可视化的整套表格工作。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SpreadsheetBench 2", "area": "专业办公", "officialUrl": "https://spreadsheetbench.github.io/", "what": "固定 V2 的 321 个任务和统一 SWE-agent 条件；独立办公产品和其他 scaffold 成绩不归给基础模型。表格可视化的正确性不等于广泛审美。", "usage": "观察中：已有稳定 JSON 和逐行日期，但当前统一条件下的基础模型覆盖不足；等待当期型号实测后再接入分数。", "limits": "观察中：已有稳定 JSON 和逐行日期，但当前统一条件下的基础模型覆盖不足；等待当期型号实测后再接入分数。", "license": "官方网站页尾声明 MIT；数据集与代码入口公开。", "attribution": "成绩由 SpreadsheetBench / Renmin University / AfterQuery 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://spreadsheetbench.github.io/", "weight": null, "status_label": "观察中" }, { "key": "vending-bench-2", "name": "Vending-Bench 2", "operator": "Andon Labs", "description": "经营模拟商店一年，处理采购、谈判、库存与定价。", "status": "observing", "brand": { "src": null, "monogram": "A", "raster": false }, "fullName": "Vending-Bench 2", "area": "专业办公", "officialUrl": "https://andonlabs.com/evals/vending-bench-2", "what": "比较同一 V2 单代理任务的期末余额，保留多次运行方差；多人竞争 Arena、不同工具和推理配置分别处理。任务中的经营成果是评测成绩，API 价格不计能力分。", "usage": "观察中：补充长期经营与电商能力的价值明确；稳定原始成绩协议、运行配置与再展示许可仍待补齐。盈利也可能来自利用模拟器，不能直接等同可信经营。", "limits": "观察中：补充长期经营与电商能力的价值明确；稳定原始成绩协议、运行配置与再展示许可仍待补齐。盈利也可能来自利用模拟器，不能直接等同可信经营。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Andon Labs 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://andonlabs.com/evals/vending-bench-2", "weight": null, "status_label": "观察中" }, { "key": "epoch-ebr", "name": "EBR-bench", "operator": "Epoch AI", "description": "在长任务中学习新规则并使用记忆", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "EBR-bench", "area": "专业办公", "officialUrl": "https://epoch.ai/benchmarks/ebr-bench", "what": "公开汇总混合了记忆压缩配置；未取得固定配置的可比成绩前仅供参考。", "usage": "仅供参考，不进入综合或分类分数。", "limits": "仅供参考，不进入综合或分类分数。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://epoch.ai/benchmarks/ebr-bench", "weight": null, "status_label": "仅供参考" }, { "key": "vals-excel", "name": "Excel · EMB", "operator": "Vals AI", "description": "表格编辑与 Excel 办公", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Excel · EMB", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/emb", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/emb", "weight": null, "status_label": "观察中" }, { "key": "vals-legal", "name": "Legal Research", "operator": "Vals AI", "description": "法律检索和论证", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "Legal Research", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/legal_research", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/legal_research", "weight": null, "status_label": "观察中" }, { "key": "vals-tax", "name": "TaxAgentBench", "operator": "Vals AI", "description": "税务专业问题", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "TaxAgentBench", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/tax_agent_bench", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/tax_agent_bench", "weight": null, "status_label": "观察中" }, { "key": "vals-medscribe", "name": "MedScribe", "operator": "Vals AI", "description": "临床记录整理与撰写", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "MedScribe", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/medscribe", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/medscribe", "weight": null, "status_label": "观察中" }, { "key": "vals-biomystery", "name": "BioMysteryBench", "operator": "Vals AI", "description": "生物学研究推理", "status": "observing", "brand": { "src": "/leaderboard-sources/vals.svg", "monogram": "V", "raster": false }, "fullName": "BioMysteryBench", "area": "专业办公", "officialUrl": "https://www.vals.ai/benchmarks/biomysterybench", "what": "专项页 Overall，与 Vals Index 综合聚合分分开。多模型回退成绩不归入单个型号。", "usage": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "limits": "观察中：专项数据的公开使用条款尚待落实；未因沿用同一网站解析器自动计分。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Vals AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.vals.ai/benchmarks/biomysterybench", "weight": null, "status_label": "观察中" }, { "key": "terminal-science", "name": "Terminal-Bench Science", "operator": "Harbor", "description": "在终端里完成科研任务", "status": "observing", "brand": { "src": null, "monogram": "H", "raster": false }, "fullName": "Terminal-Bench Science", "area": "专业办公", "officialUrl": "https://www.terminal-bench-science.ai/", "what": "区分 Claude Code、Codex 和 mini-SWE 等系统；不能把不同系统成绩直接当模型能力。", "usage": "观察中：12 个型号且混合运行系统，新前沿型号仍不完整。", "limits": "观察中：12 个型号且混合运行系统，新前沿型号仍不完整。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Harbor 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "professional", "category": "professional", "url": "https://www.terminal-bench-science.ai/", "weight": null, "status_label": "观察中" }, { "key": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "description": "事实问答，答错倒扣、拒答不扣，看模型知道什么、会不会乱编。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-Omniscience（Artificial Analysis）", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/omniscience", "what": "取 Artificial Analysis 智能指数中 Omniscience 的分数（−100 到 100）。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域，与 SimpleQA Verified 同属事实准确一类，两项合起来算一票。", "limits": "分数同时反映知识量和是否乱编，不单纯是知识量。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://artificialanalysis.ai/evaluations/omniscience", "weight": null, "status_label": "参与排名" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "description": "短问题的事实准确性", "status": "ranked", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "SimpleQA Verified", "area": "知识", "officialUrl": "https://epoch.ai/benchmarks/simple-qa-verified", "what": "只取 Epoch 自行复跑的 mean_score，采用固定推理档位；不读取 ECI 或 Best score 聚合列。", "usage": "计分：知识问答领域，与 AA-Omniscience 同属事实准确一类，两项合起来算一票。", "limits": "正式计分；同题家族共用固定预算。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://epoch.ai/benchmarks/simple-qa-verified", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "description": "读很长的文档后，回答需要综合推理的问题。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "AA-LCR（Artificial Analysis）", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "what": "取 Artificial Analysis 智能指数中 AA-LCR 的正确率。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域的长文理解一类。", "limits": "只考英文长文。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "weight": null, "status_label": "参与排名" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令", "operator": "LiveBench", "description": "Language + IF 两项均分，各占 50%。", "status": "ranked", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench · 语言与指令", "area": null, "officialUrl": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "what": "读取 Language 和 IF 的未舍入分项成绩；本来源页保留两项均分。", "usage": "计分：Language 与 IF 分别计入知识问答领域的语言理解、按要求作答两类。", "limits": "两项均分衡量组合表现，不能当成 Language 单项成绩。官方页面切换分类或是否包含微调模型时，分数与排序会随之变化。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "分项成绩由 LiveBench 发布，AIHOT 按上述两类等权汇总并排序。各项分数尺度不同，不能直接相加；参考排序采用固定参照尺度。", "group": "knowledge", "category": "knowledge", "url": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "weight": null, "status_label": "参与排名" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提", "operator": "Peter Gostev / BullshitBench", "description": "识别问题里的错误前提，观察模型会不会顺着错误继续编造。", "status": "ranked", "brand": { "src": null, "monogram": "P", "raster": false }, "fullName": "BullshitBench v2", "area": "知识", "officialUrl": "https://github.com/petergpt/bullshit-benchmark", "what": "取官方 v2 最新榜单的 green rate：所有尝试中明确识破错误前提的比例。同一型号按最高推理档位选代表配置，选择不看分数。", "usage": "计分：知识问答领域的识破错误前提一类。", "limits": "全部为错误前提题，不代表一般知识广度；由模型裁判判定。", "license": "MIT · Copyright (c) 2026 Peter Gostev；官方成绩随同仓库发布，无单独数据许可例外。", "attribution": "成绩由 Peter Gostev / BullshitBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://github.com/petergpt/bullshit-benchmark", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "description": "准确、完整且简洁地理解医疗长记录。", "status": "ranked", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "MLCR-AA", "area": "知识问答", "officialUrl": "https://artificialanalysis.ai/evaluations/mlcr-aa", "what": "25k–64k token 医疗记录；专家参考答案、三个模型裁判多数表决；每题三次，取总体 pass@1，空答或超过参考答案五倍长度失败。", "usage": "计分：长文理解；与 AA-LCR 共用一类评测。 固定十个对照型号至少测到六个，才参与计算。", "limits": "只取 mlcrOverall，包含长度失败；不替换成仅对合格长度作答的准确率。 缺测留空；回退成绩计入并标注，受影响题目过半不计。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://artificialanalysis.ai/evaluations/mlcr-aa", "weight": null, "status_label": "参与排名" }, { "key": "epoch-gpqa", "name": "GPQA Diamond", "operator": "Epoch AI", "description": "研究生水平的物理、化学与生物知识", "status": "reference_only", "brand": { "src": "/leaderboard-sources/epoch.svg", "monogram": "E", "raster": false }, "fullName": "GPQA Diamond", "area": "知识", "officialUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "what": "只取 Epoch 自行复跑的 mean_score，保留推理档位与标准误。198 道科学知识题需要推理，但不能代表全部世界知识。", "usage": "头部模型已接近满分，区分度不足，只作参考，不计分。", "limits": "头部模型已接近满分，区分度不足。", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "attribution": "成绩由 Epoch AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://epoch.ai/benchmarks/gpqa-diamond", "weight": null, "status_label": "仅供参考" }, { "key": "facts-parametric", "name": "FACTS Parametric", "operator": "Google DeepMind / Kaggle", "description": "不借助搜索工具回答世界事实问题，检验模型自身知识的准确性。", "status": "observing", "brand": { "src": null, "monogram": "G", "raster": false }, "fullName": "FACTS Parametric", "area": "知识", "officialUrl": "https://www.kaggle.com/benchmarks/google/facts-parametric/leaderboard", "what": "只考虑固定任务版本的 Score，保留公开集、私有集与置信区间供核查；不能把三个相关分数重复计票。FACTS 不在 AA v4.3 的十项底层评测中。", "usage": "观察中：官方协议与许可已确认，但没有完成题数或完整性标记；部分行区间显著宽，总分与公开集、私有集的汇总口径也存在待确认情形。等待评测方说明，不删除异常行来强行接入，不占评分预算。", "limits": "裁判与评测发起方均来自 Google，需结合其他机构证据。导出未说明部分结果是否完成同一套题，不能把局部结果当完整评测；从区间反推的样本量不是官方确认事实。", "license": "Apache 2.0 · 官方 benchmark 与 Score task 页面明确声明；Kaggle 文档提供公共榜单的免登录 JSON 下载。", "attribution": "成绩由 Google DeepMind / Kaggle 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "knowledge", "category": "knowledge", "url": "https://www.kaggle.com/benchmarks/google/facts-parametric/leaderboard", "weight": null, "status_label": "观察中" }, { "key": "arena-text", "name": "Arena Text", "operator": "LMArena", "description": "真人匿名两两比较回答，看大家更愿意读哪一个。补上选择题看不出的整体体验。", "status": "ranked", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Text Style-Controlled", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "只取官方发布快照中的 Text Overall style-controlled 主榜，不抓取 arena.ai 实时页面。", "usage": "计分：真人盲选领域，只这一项。", "limits": "人类偏好受回答风格、用户结构和提示分布影响。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "general", "category": "general", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "weight": null, "status_label": "参与排名" }, { "key": "artificial-analysis", "name": "AA Index", "operator": "Artificial Analysis", "description": "把多项当前评测放到同一套环境里复跑，得到一个综合能力分。价格只作参考，不进总分。", "status": "cross_reference", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Artificial Analysis Intelligence Index", "area": null, "officialUrl": "https://artificialanalysis.ai/", "what": "取 Artificial Analysis 公开页面上的智能指数，只作交叉核对。", "usage": "综合指数不计分，只用来核对；其中 9 项分项已分别计分，不重复计入。", "limits": "聚合指数包含多种任务和运行条件；须保留指数版本、配置与上游署名。", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "general", "category": "general", "url": "https://artificialanalysis.ai/", "weight": null, "status_label": "交叉参考" }, { "key": "livebench-general", "name": "LiveBench", "operator": "LiveBench", "description": "题库持续更换，答案可以自动核对。用来看模型换题之后还能不能站住。", "status": "cross_reference", "brand": { "src": "/leaderboard-sources/livebench.png", "monogram": "L", "raster": true }, "fullName": "LiveBench Global Average", "area": null, "officialUrl": "https://livebench.ai/", "what": "保留官方 Global Average 作整体参考；正式分数使用互不重复的专项任务。", "usage": "LiveBench 总均分只做交叉核对，不计分；它的分类成绩已分别计分。", "limits": "综合均分会稀释特定能力的变化；仅交叉参考，不重复计票。组织参与方 Abacus.AI 同时开发 Smaug 模型，其独家成绩需要其他独立来源补充。", "license": "Apache 2.0", "attribution": "成绩由 LiveBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "general", "category": "general", "url": "https://livebench.ai/", "weight": null, "status_label": "交叉参考" }, { "key": "arena-webdev", "name": "Arena WebDev", "operator": "LMArena", "description": "同样是真人盲选，比的是网页好不好用、能不能交付。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena WebDev", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "读取官方发布的 WebDev / Code Arena 快照，保留上游型号与配置。", "usage": "只作参考，不计分：部分型号经厂商自己的 Agent（如 codex-harness）参赛。", "limits": "不能把不同专用 Agent 或不同执行流程的交付表现直接归因于底层型号；待共同协议范围核实后再评估。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "general", "category": "general", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "weight": null, "status_label": "仅供参考" }, { "key": "arena-creative-writing", "name": "Arena 创作盲选", "operator": "LMArena", "description": "真人盲选故事、诗歌和其他创意表达，观察作品是否吸引读者。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Creative Writing", "area": null, "officialUrl": "https://arena.ai/leaderboard/text/creative-writing", "what": "读取官方数据集 text_style_control 的 creative_writing 分类，采用风格控制后的 Bradley–Terry 分与置信区间；不借用整体文本或职业写作分类成绩。", "usage": "写作目前不计分，只作参考。", "limits": "写作目前不计分，只作参考。", "license": "CC BY 4.0 · Arena 官方 Hugging Face leaderboard-dataset；保留署名、来源链接并说明聚合改动。", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://arena.ai/leaderboard/text/creative-writing", "weight": null, "status_label": "仅供参考" }, { "key": "eq-creative", "name": "Creative Writing v3", "operator": "EQ-Bench", "description": "短篇创作与表达", "status": "reference_only", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "Creative Writing v3", "area": null, "officialUrl": "https://eqbench.com/creative_writing.html", "what": "官方 CSV 内嵌于数据脚本，采用 Elo；Sonnet 4.6 判胜负，风格、重复率和 rubric 不重复计票。", "usage": "写作目前不计分，只作参考。", "limits": "评分来自模型裁判；写作目前不计分。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://eqbench.com/creative_writing.html", "weight": null, "status_label": "仅供参考" }, { "key": "eq-longform", "name": "Longform Writing", "operator": "EQ-Bench", "description": "长篇故事的连贯性与完整性", "status": "reference_only", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "Longform Writing", "area": null, "officialUrl": "https://eqbench.com/creative_writing_longform.html", "what": "采用 overall_score_100；八章长文的情节与表达评分，裁判偏好作为局限披露。", "usage": "写作目前不计分，只作参考。", "limits": "评分来自模型裁判；写作目前不计分。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://eqbench.com/creative_writing_longform.html", "weight": null, "status_label": "仅供参考" }, { "key": "eq-emotional-v4", "name": "EQ-Bench 4 · 情绪与人际理解", "operator": "EQ-Bench", "description": "在多轮交流中理解人物情绪、隐含需求与不同的沟通偏好。", "status": "observing", "brand": { "src": "/leaderboard-sources/eqbench.svg", "monogram": "E", "raster": true }, "fullName": "EQ-Bench 4 · 情绪与人际理解", "area": null, "officialUrl": "https://eqbench.com/", "what": "120 个模拟人物、16 轮角色扮演，三家模型裁判轮换盲评；只采用 Elo，行为风格 traits 不作越高越好的能力分。旧 v3 不重复计票。", "usage": "观察中：公开数据和许可可用，等待新一轮评测；不延长有效期来凑分类来源数。", "limits": "英文模拟交流且尚未用人类专家验证，不能等同真人满意度或中文情商；当前成绩已超 45 天窗口。", "license": "MIT · EQ-bench-site README 元数据声明", "attribution": "成绩由 EQ-Bench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://eqbench.com/", "weight": null, "status_label": "观察中" }, { "key": "lechmazur-short-story", "name": "Short-Story · 短篇小说", "operator": "Lech Mazur", "description": "把固定人物、情节与风格要求自然写进一篇完整短篇小说。", "status": "observing", "brand": { "src": null, "monogram": "L", "raster": false }, "fullName": "Short-Story · 短篇小说", "area": null, "officialUrl": "https://github.com/lechmazur/writing", "what": "600–800 词短篇、10 项创作要求；同提示、三家裁判、双向比较。采用 Thurstone 主分，BT 仅作诊断，不重复计票；多档位按固定规则选取。", "usage": "观察中：补齐正式数据复用边界并解决 README 与机器成绩不同步后再接入。", "limits": "模型裁判仍有文学偏好；部分型号未完成全部故事。短篇能力不等于长篇小说或中文文笔。", "license": "未找到明确的榜单数据复用许可；GitHub 公开不等于开放授权。", "attribution": "成绩由 Lech Mazur 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://github.com/lechmazur/writing", "weight": null, "status_label": "观察中" }, { "key": "tonebench-scripts", "name": "ToneBench · 文风与脚本", "operator": "Towards AI", "description": "按指定作者文风写视频脚本，考查开场、结构、表达与口播节奏。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "ToneBench · 文风与脚本", "area": null, "officialUrl": "https://benchmark.towardsai.com/", "what": "10 个真实脚本任务、多次生成、三家裁判；只比较相同任务、rubric、裁判和运行指纹。混合模型回退行须排除，不按每列最高成绩选配置。", "usage": "观察中：数据协议和当期覆盖较好，待确认成绩再展示边界并核验纯模型配置。", "limits": "偏向 Towards AI 的单一频道文风，精确题目和参考脚本未公开；现有第一名含 Fable 5 + Opus 4.8 回退，不能作为单型号成绩。", "license": "公开 JSON 可读取，尚未找到明确的数据聚合与再展示条款。", "attribution": "成绩由 Towards AI 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://benchmark.towardsai.com/", "weight": null, "status_label": "观察中" }, { "key": "tubelab-scriptwriting", "name": "TubeLab · 视频脚本", "operator": "TubeLab", "description": "面向不同视频频道写脚本，比较故事组织、文风、节奏与清晰度。", "status": "observing", "brand": { "src": null, "monogram": "T", "raster": false }, "fullName": "TubeLab · 视频脚本", "area": null, "officialUrl": "https://tubelab.net/benchmark/scriptwriting", "what": "统一 TubeLab Scriptwriter、12 个任务、6 项写作质量；三家模型裁判不评自己家，失败任务保留惩罚。成本和速度不进入能力。", "usage": "观察中：等待稳定原始数据、准确评测日期和复用边界，不用展示页面代替可核验成绩。", "limits": "模型受 Scriptwriter 流程影响；OpenAI 使用最低推理档位，不能外推到其它档位。视频口播不等于电影剧本。", "license": "未确认稳定成绩导出和公开聚合许可。", "attribution": "成绩由 TubeLab 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://tubelab.net/benchmark/scriptwriting", "weight": null, "status_label": "观察中" }, { "key": "ncbench-author-workflow", "name": "NC Bench · 作家工作流", "operator": "Novelcrafter", "description": "帮助作家改写、续写、总结、翻译和整理创作想法。", "status": "observing", "brand": { "src": null, "monogram": "N", "raster": false }, "fullName": "NC Bench · 作家工作流", "area": null, "officialUrl": "https://www.nc-bench.com/", "what": "只考虑独立写作和文本编辑任务的纯质量成绩；Rank Score 含价格、速度与稳定性，明确不采用。", "usage": "观察中：待核验可比的任务级原始质量成绩、真实更新日期与再展示边界。", "limits": "综合 Rank Score 混合非能力因素；不同任务覆盖不齐时不能直接用首页聚合名次。", "license": "尚未确认机器成绩与数据复用条款。", "attribution": "成绩由 Novelcrafter 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://www.nc-bench.com/", "weight": null, "status_label": "观察中" }, { "key": "design-arena-design", "name": "Design Arena · 视觉设计", "operator": "Design Arena / Intelligence", "description": "真人盲选网页、界面、图表和幻灯片的设计作品。", "status": "observing", "brand": { "src": null, "monogram": "D", "raster": false }, "fullName": "Design Arena · 视觉设计", "area": null, "officialUrl": "https://www.designarena.ai/leaderboard/website", "what": "仅比较 Models Arena 的基础模型；各设计任务共享家族预算，不把子标签当多个独立来源，也不与 codecategories 总分重复计票。", "usage": "观察中：正式 API 协议和展示条款已明确，待获批访问及逐项原始成绩核验。尚未采集、尚未计分。", "limits": "公开网页不替代 API 授权；模型、Agent 和 Builder 不能混排。", "license": "正式 API 数据允许个人及商业展示，须署名和可见官网链接；API 需申请获批。", "attribution": "成绩由 Design Arena / Intelligence 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://www.designarena.ai/leaderboard/website", "weight": null, "status_label": "观察中" }, { "key": "openvibeeval-design", "name": "OpenVibeEval · 网页作品", "operator": "OpenVibeEval", "description": "对照网页作品、可访问性和盲选偏好，观察不同工具下的设计表现。", "status": "observing", "brand": { "src": null, "monogram": "O", "raster": false }, "fullName": "OpenVibeEval · 网页作品", "area": null, "officialUrl": "https://openvibeeval.com/", "what": "模型与 harness 分开保存；可访问性分不是审美分，需要核验独立盲选成绩与同运行条件才能采用。", "usage": "观察中：需补齐同条件、足量盲测的原始偏好数据及发布时间。", "limits": "模型混用不同运行工具；主表是可访问性与任务成果，不是统一审美投票。", "license": "未确认稳定成绩导出与公开聚合许可。", "attribution": "成绩由 OpenVibeEval 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://openvibeeval.com/", "weight": null, "status_label": "观察中" }, { "key": "svgbench-arena", "name": "SVGBench · 矢量图盲选", "operator": "SVGBench", "description": "按相同提示盲选 SVG，考查视觉清晰度与图形表现。", "status": "observing", "brand": { "src": null, "monogram": "S", "raster": false }, "fullName": "SVGBench · 矢量图盲选", "area": null, "officialUrl": "https://svgbench.ai/leaderboard", "what": "需要完整、已达到提示覆盖门槛的盲测分数；注册模型数量不等于有效上榜型号数。与其它 SVG 评测共用相关任务家族。", "usage": "观察中：等待稳定原始成绩、可验证更新时间与再展示边界。", "limits": "网页统计不能证明每个新型号都有足量评测；没有原始表和时间时不声称已覆盖。", "license": "未找到明确的成绩复用许可。", "attribution": "成绩由 SVGBench 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://svgbench.ai/leaderboard", "weight": null, "status_label": "观察中" }, { "key": "rapidata-svg", "name": "Rapidata SVG", "operator": "Rapidata", "description": "真人比较模型生成的矢量图，直接评价视觉偏好。", "status": "observing", "brand": { "src": "/leaderboard-sources/rapidata.svg", "monogram": "R", "raster": false }, "fullName": "Rapidata SVG · Human Preference", "area": null, "officialUrl": "https://www.benchmark.ai/", "what": "只采用同一静态 SVG 任务的 Preference Elo；Alignment、Coherence 和 Overall 不重复计票。以官方数据集固定版本的公开成绩表为准，可追溯原始人类对战。", "usage": "观察中：官方数据与许可可用，但三条 DeepSeek 滚动别名缺少固定版本证据，整张成绩暂不采纳。等待评测方补充运行版本；没有削减现役来源预算。", "limits": "SVG 绘图不能代表全部界面设计。公开原始数据不含模型版本、生成日期或 API fingerprint，无法判断滚动接口是否换版；不能用榜单更新日猜模型版本。", "license": "CC BY 4.0 · Rapidata/svg-benchmark；保留官方署名与数据链接。", "attribution": "成绩由 Rapidata 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "writing", "category": "writing", "url": "https://www.benchmark.ai/", "weight": null, "status_label": "观察中" }, { "key": "arena-vision", "name": "Arena Vision", "operator": "LMArena", "description": "真人盲选带图片的回答，观察视觉理解的实际体验。", "status": "reference_only", "brand": { "src": "/leaderboard-sources/arena.svg", "monogram": "L", "raster": false }, "fullName": "Arena Vision Style-Controlled", "area": null, "officialUrl": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "what": "只读取官方数据集 vision_style_control 的 overall；不含图像生成或图像编辑。", "usage": "视觉目前不计分，只作参考。", "limits": "视觉目前不计分，只作参考。", "license": "CC BY 4.0", "attribution": "成绩由 LMArena 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "vision", "category": "vision", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "weight": null, "status_label": "仅供参考" }, { "key": "roboflow-vision", "name": "Roboflow Vision Evals", "operator": "Roboflow", "description": "读图、计数、OCR 与图像信息提取", "status": "observing", "brand": { "src": null, "monogram": "R", "raster": false }, "fullName": "Roboflow Vision Evals", "area": null, "officialUrl": "https://playground.roboflow.com/evals", "what": "六项任务共同预算；区分严格准确率与模型裁判分，不混用三次运行半极差与 95% 区间。", "usage": "观察中：新榜数据接口返回限流；旧开源仓库不含这一轮新型号，不用旧数据冒充新榜。", "limits": "观察中：新榜数据接口返回限流；旧开源仓库不含这一轮新型号，不用旧数据冒充新榜。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 Roboflow 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "vision", "category": "vision", "url": "https://playground.roboflow.com/evals", "weight": null, "status_label": "观察中" }, { "key": "artificial-analysis-multilingual", "name": "AA 中文／多语言", "operator": "Artificial Analysis", "description": "单独看中文和多语言表现，不再把英文综合能力重复算一次。", "status": "observing", "brand": { "src": "/leaderboard-sources/artificial-analysis-purple-2026.svg", "monogram": "A", "raster": false }, "fullName": "Artificial Analysis Multilingual Index", "area": null, "officialUrl": "https://artificialanalysis.ai/models/multilingual", "what": "只取官方多语言指数字段，不拆单语子分再投一次。", "usage": "中文与多语言目前不计分。", "limits": "中文与多语言目前不计分。", "license": "API 数据展示或分享须显著署名；再分发权与定制条款须另行取得并遵守 Terms of Use", "attribution": "成绩由 Artificial Analysis 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "multilingual", "category": "multilingual", "url": "https://artificialanalysis.ai/models/multilingual", "weight": null, "status_label": "观察中" }, { "key": "rws-mgate", "name": "RWS M-GATE", "operator": "RWS", "description": "含中文的 30 种语言，考查语法与翻译", "status": "observing", "brand": { "src": null, "monogram": "R", "raster": false }, "fullName": "RWS M-GATE", "area": null, "officialUrl": "https://www.rws.com/artificial-intelligence/train-ai-data-services/research/m-gate/leaderboard/", "what": "语法 MCC 与翻译分别保存；分词效率不进入能力分数。", "usage": "观察中：浏览器可见新模型，但机器接口返回空内容；缺少可验证的成绩发布时间与稳定导出。", "limits": "观察中：浏览器可见新模型，但机器接口返回空内容；缺少可验证的成绩发布时间与稳定导出。", "license": "待确认榜单数据使用边界", "attribution": "成绩由 RWS 发布。各项评测的原始分数尺度不同，不能直接相加；AIHOT 按公开规则把它们换算到同一条能力刻度上。", "group": "multilingual", "category": "multilingual", "url": "https://www.rws.com/artificial-intelligence/train-ai-data-services/research/m-gate/leaderboard/", "weight": null, "status_label": "观察中" } ], "raw_proofs": [ { "name": "professional", "url": "https://aihot.news/api/site/leaderboard/boards/professional", "status": 200, "fetched_at": "2026-10-09T04:16:18.572Z", "sha256": "7e2d511b3541ee86ea07eab1c6623fffbf510fff42d9e6c77834d680227e6555" }, { "name": "overall", "url": "https://aihot.news/api/site/leaderboard/boards/overall", "status": 200, "fetched_at": "2026-10-09T04:16:18.729Z", "sha256": "13b4882792a1f929253dd26f618c575fce0d66caf9302ca8b46b51a384b5dee6" }, { "name": "coding", "url": "https://aihot.news/api/site/leaderboard/boards/coding", "status": 200, "fetched_at": "2026-10-09T04:16:18.776Z", "sha256": "79dc9471e6d80ca71955c26761e0cd66bc2ef1fc4b080480c6a93b39248224a4" }, { "name": "reasoning", "url": "https://aihot.news/api/site/leaderboard/boards/reasoning", "status": 200, "fetched_at": "2026-10-09T04:16:18.885Z", "sha256": "45c4244ad9c0df38bc9b1e6b7ba61e3bad625be6cae3668197711de93694b498" }, { "name": "knowledge", "url": "https://aihot.news/api/site/leaderboard/boards/knowledge", "status": 200, "fetched_at": "2026-10-09T04:16:19.967Z", "sha256": "bce9dfe5e63c8c73875cbeec24a1033df72490f56bf0e2ee7b0864cd66ddb663" }, { "name": "rules", "url": "https://aihot.news/api/site/leaderboard/rules", "status": 200, "fetched_at": "2026-10-09T04:16:20.554Z", "sha256": "2b481a806a337022e683eec3f2ce3dd9eb0f81613e0609b02532f81fad80364d" }, { "name": "sources", "url": "https://aihot.news/api/site/leaderboard/sources", "status": 200, "fetched_at": "2026-10-09T04:16:21.001Z", "sha256": "d61e47cb6b5cd0f38b9608ba6b0cfab80aeeda39f92c0432df056fe30a753e59" }, { "name": "registry", "url": "https://api.github.com/repos/KKKKhazix/AIHOT/contents/packages/backend/src/leaderboard/source-registry.json?ref=3343fe2b20db4be7269113752d82d3992fc52b6b", "status": 200, "fetched_at": "2026-10-09T04:16:21.442Z", "sha256": "6cccc872adce3b50b03c10b416925a09a6dab517b806e41454e684a4f3a2a016" }, { "name": "source-vals-terminal-bench-4", "url": "https://aihot.news/api/site/leaderboard/sources/vals-terminal-bench-4", "status": 200, "fetched_at": "2026-10-09T04:16:22.525Z", "sha256": "95515dfd9d95e0e61a8ee1f269f541e9bb54f7c4d70f507f695a1cebf4c35eac" }, { "name": "source-artificial-analysis-terminal-bench", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-terminal-bench", "status": 200, "fetched_at": "2026-10-09T04:16:22.541Z", "sha256": "6d1cdc7833c9920be8f58be74953e0ca6ccacf5dd64a5932515dd402fd045fdd" }, { "name": "source-livebench-coding", "url": "https://aihot.news/api/site/leaderboard/sources/livebench-coding", "status": 200, "fetched_at": "2026-10-09T04:16:22.626Z", "sha256": "d7317af231758a917f08a8a6d6a576b1c7010cfba457a9a0d8227629adf9ef73" }, { "name": "source-artificial-analysis-scicode", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-scicode", "status": 200, "fetched_at": "2026-10-09T04:16:23.657Z", "sha256": "5117f392bae92f1ad69ea7a926410da267ae65ddbab106996de2578c7dbd388b" }, { "name": "source-vals-migration", "url": "https://aihot.news/api/site/leaderboard/sources/vals-migration", "status": 200, "fetched_at": "2026-10-09T04:16:24.122Z", "sha256": "38b1c3f20fb3e74558cf802afe31e0c3a207c84f5508e77e7c9b6a2fdf6faf58" }, { "name": "source-vals-vcb-1-100", "url": "https://aihot.news/api/site/leaderboard/sources/vals-vcb-1-100", "status": 200, "fetched_at": "2026-10-09T04:16:24.798Z", "sha256": "31c0954bd1dd018d0aadcebb2dfe690ea11cb8750888d7cd1db6ace4fd0bd00c" }, { "name": "source-vals-vibe-code", "url": "https://aihot.news/api/site/leaderboard/sources/vals-vibe-code", "status": 200, "fetched_at": "2026-10-09T04:16:24.848Z", "sha256": "e4ee92d2c2d1a4a9ff74b2a1f40a829436ea3f56d3ec21cf4784c83ae20a097a" }, { "name": "source-vals-program", "url": "https://aihot.news/api/site/leaderboard/sources/vals-program", "status": 200, "fetched_at": "2026-10-09T04:16:24.872Z", "sha256": "550995af05058e0f287b250f480727389f18ff738094fb35e6b95b4ec240914d" }, { "name": "source-taptap-maker", "url": "https://aihot.news/api/site/leaderboard/sources/taptap-maker", "status": 200, "fetched_at": "2026-10-09T04:16:25.948Z", "sha256": "391214261282b10d41b2771d1a8d3d4e32d893f09de0d493c2e8ab96a1900b1a" }, { "name": "source-epoch-mirrorcode", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-mirrorcode", "status": 200, "fetched_at": "2026-10-09T04:16:25.999Z", "sha256": "2748c5df342f7e1cea9b27b9e87322c0e7bec609f140233746ac703bb2596724" }, { "name": "source-deepswe-v1-1", "url": "https://aihot.news/api/site/leaderboard/sources/deepswe-v1-1", "status": 200, "fetched_at": "2026-10-09T04:16:26.002Z", "sha256": "452642759b7e0afbe094c262062e1ea51943ddc30c319f3310b49fdd38173097" }, { "name": "source-vals-cyber-patch", "url": "https://aihot.news/api/site/leaderboard/sources/vals-cyber-patch", "status": 200, "fetched_at": "2026-10-09T04:16:26.013Z", "sha256": "9149183c9d96d1e15fd0f9e2f377b41a8ae9065b1b73fa724286ed9fade26be1" }, { "name": "source-terminal-bench-4", "url": "https://aihot.news/api/site/leaderboard/sources/terminal-bench-4", "status": 200, "fetched_at": "2026-10-09T04:16:27.064Z", "sha256": "59d354b98e3ce683d780baca17cda2f894f469e35f0894df2faf1924f35df0c2" }, { "name": "source-hyper-tau", "url": "https://aihot.news/api/site/leaderboard/sources/hyper-tau", "status": 200, "fetched_at": "2026-10-09T04:16:27.147Z", "sha256": "8798d8d5af24f72fabce16bb33e3a2a323c06044a793b9dfdb8ce554c3bcb066" }, { "name": "source-swe-rebench", "url": "https://aihot.news/api/site/leaderboard/sources/swe-rebench", "status": 200, "fetched_at": "2026-10-09T04:16:27.827Z", "sha256": "9518e9886592df5493d14619d7fc953c82f41f5a62c1af17e3e5d666b182bbc6" }, { "name": "source-lhtb", "url": "https://aihot.news/api/site/leaderboard/sources/lhtb", "status": 200, "fetched_at": "2026-10-09T04:16:27.842Z", "sha256": "8110b045fee89580197e0aef1974051fa492ba643d6273b13916d6767be8dcfa" }, { "name": "source-frontiercode", "url": "https://aihot.news/api/site/leaderboard/sources/frontiercode", "status": 200, "fetched_at": "2026-10-09T04:16:28.291Z", "sha256": "17695d55bad5c6da54dbab14705cab33657434d0a9a0a368aefad3a9cd054a19" }, { "name": "source-weirdml", "url": "https://aihot.news/api/site/leaderboard/sources/weirdml", "status": 200, "fetched_at": "2026-10-09T04:16:28.965Z", "sha256": "01191bc9363c1e228945223307bfe15c5f89293e73c41891813d4e6526e4c94d" }, { "name": "source-frontierswe-v2", "url": "https://aihot.news/api/site/leaderboard/sources/frontierswe-v2", "status": 200, "fetched_at": "2026-10-09T04:16:29.049Z", "sha256": "011f7bbb1cbc9b9839ce3851e41cd5dfde127cf98c02ba23ccc24aa79d4f0332" }, { "name": "source-artificial-analysis-hle", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-hle", "status": 200, "fetched_at": "2026-10-09T04:16:29.244Z", "sha256": "b5cedea42d535ea1a6c4a51de810f351621871b44036a361027ba612541f73f2" }, { "name": "source-epoch-frontiermath", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-frontiermath", "status": 200, "fetched_at": "2026-10-09T04:16:30.298Z", "sha256": "a1af704aefe185a3253b0c5d37c7a901effa9538feab3e0de97459f1356e4486" }, { "name": "source-epoch-chess", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-chess", "status": 200, "fetched_at": "2026-10-09T04:16:30.324Z", "sha256": "2af49874eff7498cb851f5dd91190877a4ef92af7f18ebb12cd4e78311a680fe" }, { "name": "source-livebench-reasoning", "url": "https://aihot.news/api/site/leaderboard/sources/livebench-reasoning", "status": 200, "fetched_at": "2026-10-09T04:16:30.441Z", "sha256": "399a70109e6fcab8b25edd795c8a6fc3e919d26c04864552ffd374da9e650c8d" }, { "name": "source-artificial-analysis-critpt", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-critpt", "status": 200, "fetched_at": "2026-10-09T04:16:30.448Z", "sha256": "a7647288ba1294d2586e4648543bb735bfb88e532610f77cfd32076d5018b08f" }, { "name": "source-epoch-mystery", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-mystery", "status": 200, "fetched_at": "2026-10-09T04:16:31.523Z", "sha256": "339e4f5d5bc142342d3ccc5721349ba8f0306c2413e7b7528c5544a7b026ba20" }, { "name": "source-vals-mysterymechanism", "url": "https://aihot.news/api/site/leaderboard/sources/vals-mysterymechanism", "status": 200, "fetched_at": "2026-10-09T04:16:31.523Z", "sha256": "e06502f24240ee9f603e9c0bca3a30da12b04f9212edcf9b724218f844c51c3b" }, { "name": "source-artificial-analysis-terminal-bench-science", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-terminal-bench-science", "status": 200, "fetched_at": "2026-10-09T04:16:31.590Z", "sha256": "0bbd89bc926e2bacc664c6a88a8e2c3643b1225216a8d86432fe8cf276f279b7" }, { "name": "source-vals-terminal-bench-science", "url": "https://aihot.news/api/site/leaderboard/sources/vals-terminal-bench-science", "status": 200, "fetched_at": "2026-10-09T04:16:32.184Z", "sha256": "c965af80207391790a5404a75c3d760858482b9ebef423a1ebaebd7c395814f2" }, { "name": "source-epoch-frontiermath-tier4", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-frontiermath-tier4", "status": 200, "fetched_at": "2026-10-09T04:16:33.202Z", "sha256": "3acf961026e764d487072ed3b00b88236524ef5d885b2128c78c0db7ba0501ca" }, { "name": "source-matharena-lean", "url": "https://aihot.news/api/site/leaderboard/sources/matharena-lean", "status": 200, "fetched_at": "2026-10-09T04:16:33.239Z", "sha256": "261ddf9a06786909e277e72b20a4cd2da3cc580c7feb6e7e88ee2bded294ad71" }, { "name": "source-hle", "url": "https://aihot.news/api/site/leaderboard/sources/hle", "status": 200, "fetched_at": "2026-10-09T04:16:33.260Z", "sha256": "fb4a760d71ce26e3982b4ca3efe1d34dab7855b0f02a729d26ebc3e099b1d264" }, { "name": "source-simplebench", "url": "https://aihot.news/api/site/leaderboard/sources/simplebench", "status": 200, "fetched_at": "2026-10-09T04:16:33.411Z", "sha256": "834b742060314c8e44023b379e06c099b663077ec12a44697ec6f9dc5039b0ed" }, { "name": "source-matharena-broken", "url": "https://aihot.news/api/site/leaderboard/sources/matharena-broken", "status": 200, "fetched_at": "2026-10-09T04:16:33.830Z", "sha256": "716b30fc58aaee5e3dd7bf93173b48a6637dbf006bca14168754804a5b657e07" }, { "name": "source-arc-agi-3", "url": "https://aihot.news/api/site/leaderboard/sources/arc-agi-3", "status": 200, "fetched_at": "2026-10-09T04:16:33.951Z", "sha256": "9c887145ea500196e13bf1233a9f3cb5ad10ee584eb48ceb9fa8e38d3c227f0c" }, { "name": "source-arc-agi-2", "url": "https://aihot.news/api/site/leaderboard/sources/arc-agi-2", "status": 200, "fetched_at": "2026-10-09T04:16:34.624Z", "sha256": "dffb8696de48377376b638fe85053dd02a5d016a84efc415a0285de84dd1315b" }, { "name": "source-matharena-math", "url": "https://aihot.news/api/site/leaderboard/sources/matharena-math", "status": 200, "fetched_at": "2026-10-09T04:16:34.625Z", "sha256": "8d96ab3d7beede3c965d53b16fdbf803d5ad448b842648255f1e1603f9993b74" }, { "name": "source-artificial-analysis-automationbench", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-automationbench", "status": 200, "fetched_at": "2026-10-09T04:16:35.137Z", "sha256": "2a61a6bffa5e435f93d0936b7525eb25188f155440cbd4a800865de82b9ebe44" }, { "name": "source-vals-finance-agent", "url": "https://aihot.news/api/site/leaderboard/sources/vals-finance-agent", "status": 200, "fetched_at": "2026-10-09T04:16:35.806Z", "sha256": "01603d09568388fb0dc82a5535997588dd074a79857b0c74ec2257ddc67f0eb0" }, { "name": "source-artificial-analysis-gdpval", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-gdpval", "status": 200, "fetched_at": "2026-10-09T04:16:36.090Z", "sha256": "3495d6a23ca6755784e7f89d0085ef9ed027b926d4d783d7dde2128e005dd796" }, { "name": "source-artificial-analysis-briefcase", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-briefcase", "status": 200, "fetched_at": "2026-10-09T04:16:36.387Z", "sha256": "db67815e95c00eacf982af1fc419448530c4d3a9e31333b69b02e94b3f71ecf6" }, { "name": "source-vals-harvey-lab", "url": "https://aihot.news/api/site/leaderboard/sources/vals-harvey-lab", "status": 200, "fetched_at": "2026-10-09T04:16:36.979Z", "sha256": "5c96748c239a1a5270c2a3cc39f93f269500c3549bb7d95899051d55a820709d" }, { "name": "source-mercor-apex-agents", "url": "https://aihot.news/api/site/leaderboard/sources/mercor-apex-agents", "status": 200, "fetched_at": "2026-10-09T04:16:37.467Z", "sha256": "d8c9d315f2c7df2e9fffe9421b5de5b407b031be0f58746222fe8e258ac5e002" }, { "name": "source-livebench-data-analysis", "url": "https://aihot.news/api/site/leaderboard/sources/livebench-data-analysis", "status": 200, "fetched_at": "2026-10-09T04:16:37.513Z", "sha256": "b0e9b3f2cf328fe23f65727c933ef652a35c96d284603677e98180618e173a2b" }, { "name": "source-mercor-tau3-banking", "url": "https://aihot.news/api/site/leaderboard/sources/mercor-tau3-banking", "status": 200, "fetched_at": "2026-10-09T04:16:37.514Z", "sha256": "487a9a82bf3e9d7362200df34068102f121b4ee541f6cc7222d480d7ebfb1fbc" }, { "name": "source-vals-medcode", "url": "https://aihot.news/api/site/leaderboard/sources/vals-medcode", "status": 200, "fetched_at": "2026-10-09T04:16:37.958Z", "sha256": "3481be7ec71c5f6e46fb8ec17dd8a847ddcf16de6f88a0e936879b26238e1234" }, { "name": "source-zapier-automation", "url": "https://aihot.news/api/site/leaderboard/sources/zapier-automation", "status": 200, "fetched_at": "2026-10-09T04:16:38.090Z", "sha256": "e2367467ffb2f661e60c3e06276c7c655a7a021867cd1d14c577254305d2584b" }, { "name": "source-tau-banking", "url": "https://aihot.news/api/site/leaderboard/sources/tau-banking", "status": 200, "fetched_at": "2026-10-09T04:16:38.652Z", "sha256": "d0f57bb48f18b0c2edadbfdd5a60bf34455f4f2055d952485e6836ffbad263d3" }, { "name": "source-artificial-analysis-analyst-agent", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-analyst-agent", "status": 200, "fetched_at": "2026-10-09T04:16:42.293Z", "sha256": "54faff275af295bd61f983ebfa662a220be25a917a07f4dee4b7cf3415088a58" }, { "name": "source-mcp-atlas", "url": "https://aihot.news/api/site/leaderboard/sources/mcp-atlas", "status": 200, "fetched_at": "2026-10-09T04:16:42.834Z", "sha256": "584dd6e616a65a5757266c0b56678220a6c570b297ddc03102c10fda49a8bb52" }, { "name": "source-financebenchmark", "url": "https://aihot.news/api/site/leaderboard/sources/financebenchmark", "status": 200, "fetched_at": "2026-10-09T04:16:43.457Z", "sha256": "a1bc8d5faca2ae3ad2a6b9931aae296b611a3ccc42d96e32714880c2e195ed41" }, { "name": "source-officeqa-pro", "url": "https://aihot.news/api/site/leaderboard/sources/officeqa-pro", "status": 200, "fetched_at": "2026-10-09T04:16:43.760Z", "sha256": "4602aac2d12a650363ccfa0e75800d3a383d27974a903fde5ae0a927ce41ef72" }, { "name": "source-prbench-legal", "url": "https://aihot.news/api/site/leaderboard/sources/prbench-legal", "status": 200, "fetched_at": "2026-10-09T04:16:43.994Z", "sha256": "cd7d77c12080707fb902247b790cca043e2edd8595f90e105d19689ca923c109" }, { "name": "source-prbench-finance", "url": "https://aihot.news/api/site/leaderboard/sources/prbench-finance", "status": 200, "fetched_at": "2026-10-09T04:16:44.446Z", "sha256": "53302ba93218792f8672915f95f3abdb30911eed8946d76e0cafae9c41a18681" }, { "name": "source-epoch-ebr", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-ebr", "status": 200, "fetched_at": "2026-10-09T04:16:45.064Z", "sha256": "2e3f2109d9a943974751f5cd09407ea1cd97e348bb53ac3c2b191ac5777627ad" }, { "name": "source-vending-bench-2", "url": "https://aihot.news/api/site/leaderboard/sources/vending-bench-2", "status": 200, "fetched_at": "2026-10-09T04:16:45.107Z", "sha256": "0956e5cddea195db5c4636b3ac9f6ec6548280cc9d6a38d4ba9895d2d734ccc1" }, { "name": "source-spreadsheetbench-2", "url": "https://aihot.news/api/site/leaderboard/sources/spreadsheetbench-2", "status": 200, "fetched_at": "2026-10-09T04:16:45.863Z", "sha256": "bbdef4264e8a7df75781952c0d79c714b453ff8a4c736e5183d6b8cdf4b96a2a" }, { "name": "source-vals-legal", "url": "https://aihot.news/api/site/leaderboard/sources/vals-legal", "status": 200, "fetched_at": "2026-10-09T04:16:46.422Z", "sha256": "de247e4db9c0b8d8765bafc5d33ed52fef0b648d53fc560f9949d24aa457c002" }, { "name": "source-vals-excel", "url": "https://aihot.news/api/site/leaderboard/sources/vals-excel", "status": 200, "fetched_at": "2026-10-09T04:16:46.982Z", "sha256": "4359113980a928dca952587534033c5a46b0ca0c6bd1832980f08bc617b018ae" }, { "name": "source-vals-medscribe", "url": "https://aihot.news/api/site/leaderboard/sources/vals-medscribe", "status": 200, "fetched_at": "2026-10-09T04:16:46.988Z", "sha256": "32a4a763ed30d6e82772f56a56fdec515195f05f9dbec0a9bef1bf4bdf8e680f" }, { "name": "source-vals-tax", "url": "https://aihot.news/api/site/leaderboard/sources/vals-tax", "status": 200, "fetched_at": "2026-10-09T04:16:47.060Z", "sha256": "347f32534db662d76df469eb7d6ee02d13791a69ccb9a984e2ba471575626e07" }, { "name": "source-artificial-analysis-omniscience", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-omniscience", "status": 200, "fetched_at": "2026-10-09T04:16:48.239Z", "sha256": "fd3712c89bcef3eab63f79bb45dce9233f63760d1ec89f19cebba75912482aa5" }, { "name": "source-epoch-simpleqa", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-simpleqa", "status": 200, "fetched_at": "2026-10-09T04:16:48.291Z", "sha256": "b546145b0db24448a4f55a12dc47f550f54ba146fc7ce4a333012822ae9406bf" }, { "name": "source-terminal-science", "url": "https://aihot.news/api/site/leaderboard/sources/terminal-science", "status": 200, "fetched_at": "2026-10-09T04:16:48.778Z", "sha256": "3ad50f5d1cbe2186afe726e19a702dda2d2d819ef36afb80aec711333d98ed97" }, { "name": "source-vals-biomystery", "url": "https://aihot.news/api/site/leaderboard/sources/vals-biomystery", "status": 200, "fetched_at": "2026-10-09T04:16:48.806Z", "sha256": "f3117819d8e736201ae52d702376d42a828f3e42075298c8c9aebe26b70148a1" }, { "name": "source-artificial-analysis-mlcr", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-mlcr", "status": 200, "fetched_at": "2026-10-09T04:16:49.870Z", "sha256": "4256a26af7c919261efebf8f41fda07234392852f8108ea79040a5111c9815c9" }, { "name": "source-livebench-writing", "url": "https://aihot.news/api/site/leaderboard/sources/livebench-writing", "status": 200, "fetched_at": "2026-10-09T04:16:49.879Z", "sha256": "39b967da193b0a2ce3107bb44e8a4774c708193f853a2d972978da0b395fb07d" }, { "name": "source-artificial-analysis-lcr", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-lcr", "status": 200, "fetched_at": "2026-10-09T04:16:49.926Z", "sha256": "8133491658d9c61dec11e53f952a78e2b994ecdfed46451efe25a00126cecd26" }, { "name": "source-bullshitbench-v2", "url": "https://aihot.news/api/site/leaderboard/sources/bullshitbench-v2", "status": 200, "fetched_at": "2026-10-09T04:16:50.612Z", "sha256": "e9661ec72f5e6f2ecd007e41dc0de64bf1483707dc1c955de92e549fe9c154ed" }, { "name": "source-arena-text", "url": "https://aihot.news/api/site/leaderboard/sources/arena-text", "status": 200, "fetched_at": "2026-10-09T04:16:51.113Z", "sha256": "f90033508e1c27cd414c82c9299cb60fe4f4ad7689bd0d2abf965626245ae030" }, { "name": "source-facts-parametric", "url": "https://aihot.news/api/site/leaderboard/sources/facts-parametric", "status": 200, "fetched_at": "2026-10-09T04:16:51.158Z", "sha256": "e0957458922522a107edb7ea4253985c58dd7eae7397789e25847755dd1594d7" }, { "name": "source-epoch-gpqa", "url": "https://aihot.news/api/site/leaderboard/sources/epoch-gpqa", "status": 200, "fetched_at": "2026-10-09T04:16:51.686Z", "sha256": "4c556cee3478b2b82715ba7bd2011cc5ffc8690b7b7cabb1eb236727842ee54a" }, { "name": "source-artificial-analysis", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis", "status": 200, "fetched_at": "2026-10-09T04:16:51.858Z", "sha256": "520fc69ce5a7ea1481fed5b78b32a8bdffceee847dd8172adf7dfbc325c70ff3" }, { "name": "source-arena-webdev", "url": "https://aihot.news/api/site/leaderboard/sources/arena-webdev", "status": 200, "fetched_at": "2026-10-09T04:16:52.388Z", "sha256": "e29708dbbabbc861a97303c6650e920ebe5f5df386839d716389271129d96b52" }, { "name": "source-livebench-general", "url": "https://aihot.news/api/site/leaderboard/sources/livebench-general", "status": 200, "fetched_at": "2026-10-09T04:16:52.925Z", "sha256": "453247a04756acd8d02d8715631cfe8b135c71aa94597267aa13255f58bb9260" }, { "name": "source-arena-creative-writing", "url": "https://aihot.news/api/site/leaderboard/sources/arena-creative-writing", "status": 200, "fetched_at": "2026-10-09T04:16:53.002Z", "sha256": "33e844a55def68fcf5a407cfd7bc4f8f4cdcd145f3c1de1408f0973a7249d5bb" }, { "name": "source-eq-creative", "url": "https://aihot.news/api/site/leaderboard/sources/eq-creative", "status": 200, "fetched_at": "2026-10-09T04:16:53.018Z", "sha256": "f89852d04fbef327f9c6fc8da03ff0e5313849434d65bf1271147de1a519e0a6" }, { "name": "source-tonebench-scripts", "url": "https://aihot.news/api/site/leaderboard/sources/tonebench-scripts", "status": 200, "fetched_at": "2026-10-09T04:16:53.576Z", "sha256": "49a1b5768ca0155d4d6c5480e17062bf7e8bb71552747dcc231db15c4c6d389b" }, { "name": "source-eq-longform", "url": "https://aihot.news/api/site/leaderboard/sources/eq-longform", "status": 200, "fetched_at": "2026-10-09T04:16:54.092Z", "sha256": "579bff4b3c911319468958428a17a7fd0d74189907898aa7cc01ce785a6b1634" }, { "name": "source-lechmazur-short-story", "url": "https://aihot.news/api/site/leaderboard/sources/lechmazur-short-story", "status": 200, "fetched_at": "2026-10-09T04:16:54.156Z", "sha256": "e86a2e70446c68376d21fffc229ea916ec4717310a5b5da359a3767885c4bb49" }, { "name": "source-eq-emotional-v4", "url": "https://aihot.news/api/site/leaderboard/sources/eq-emotional-v4", "status": 200, "fetched_at": "2026-10-09T04:16:54.773Z", "sha256": "de84208ec3713d81821a720531cccebcb9da24a40c475c0ac0a4ca3c8a25eec2" }, { "name": "source-ncbench-author-workflow", "url": "https://aihot.news/api/site/leaderboard/sources/ncbench-author-workflow", "status": 200, "fetched_at": "2026-10-09T04:16:55.355Z", "sha256": "db1b940e5b463a365f51109ae1ffa59878fc5b97630ffffdf1e9667ec9a0b318" }, { "name": "source-tubelab-scriptwriting", "url": "https://aihot.news/api/site/leaderboard/sources/tubelab-scriptwriting", "status": 200, "fetched_at": "2026-10-09T04:16:55.872Z", "sha256": "9f9aa506bd53da51844e19f7d7ceb5c4c6bd20fd09062ca19d5b314d60c55a16" }, { "name": "source-openvibeeval-design", "url": "https://aihot.news/api/site/leaderboard/sources/openvibeeval-design", "status": 200, "fetched_at": "2026-10-09T04:16:55.942Z", "sha256": "a8916c1554d461bf54e74bde83610bc590160802959d4aee030fcae98497f1c9" }, { "name": "source-design-arena-design", "url": "https://aihot.news/api/site/leaderboard/sources/design-arena-design", "status": 200, "fetched_at": "2026-10-09T04:16:55.997Z", "sha256": "18ff292465b1cfc453d8b1f33772069209fd8ca99f6c7aaca5204bd86eb29344" }, { "name": "source-svgbench-arena", "url": "https://aihot.news/api/site/leaderboard/sources/svgbench-arena", "status": 200, "fetched_at": "2026-10-09T04:16:56.444Z", "sha256": "2759a3797ca40680fbf2247509654cc5863d18b984b3e989f0f48bcd288d2307" }, { "name": "source-rapidata-svg", "url": "https://aihot.news/api/site/leaderboard/sources/rapidata-svg", "status": 200, "fetched_at": "2026-10-09T04:16:56.552Z", "sha256": "c77497181c7016fa278754c4d3e75fecca068af1bdf3028adfef330514ac203f" }, { "name": "source-arena-vision", "url": "https://aihot.news/api/site/leaderboard/sources/arena-vision", "status": 200, "fetched_at": "2026-10-09T04:16:57.158Z", "sha256": "e2c04431cc65122a314ad397c2ac04a8ffcfa6adfefa1125db0d0da608245018" }, { "name": "source-roboflow-vision", "url": "https://aihot.news/api/site/leaderboard/sources/roboflow-vision", "status": 200, "fetched_at": "2026-10-09T04:16:57.218Z", "sha256": "e85bc4e628aabd3af21fb85e7bc9a020ad8a4986c526cc74d54e667aa5ac4e0b" }, { "name": "source-rws-mgate", "url": "https://aihot.news/api/site/leaderboard/sources/rws-mgate", "status": 200, "fetched_at": "2026-10-09T04:16:58.346Z", "sha256": "f003bca898c35281969041d6b60f7907e6d27b9e5cca8d657640ae44883e5221" }, { "name": "source-artificial-analysis-multilingual", "url": "https://aihot.news/api/site/leaderboard/sources/artificial-analysis-multilingual", "status": 200, "fetched_at": "2026-10-09T04:16:58.346Z", "sha256": "eb1f692624a6479e3efd261a6373a5e3ab6657294256df6c46d09c52a18a7b33" } ] }, "rules": { "budgets": [ { "key": "coding", "name": "编程", "weight": 0.4, "sources": [ "Terminal-Bench 4.0 · Artificial Analysis", "Terminal-Bench 4.0 · Vals AI", "LiveBench Coding · LiveBench", "SciCode · Artificial Analysis", "Code Migration · Vals AI", "ProgramBench · Vals AI", "Vibe Code Bench v1.1 · Vals AI", "Vibe Code Bench 1–100 · Vals AI", "CyberBench Patch v1.1 · Vals AI" ] }, { "key": "reasoning", "name": "科研", "weight": 0.2, "sources": [ "Humanity's Last Exam · Artificial Analysis", "FrontierMath T1–3 · Epoch AI", "LiveBench Reasoning · LiveBench", "CritPt · Artificial Analysis", "Chess Puzzles · Epoch AI", "Mystery Game Puzzles · Epoch AI", "MysteryMechanism · Vals AI", "Terminal-Bench Science · Vals AI", "Terminal-Bench Science · Artificial Analysis" ] }, { "key": "professional", "name": "专业办公", "weight": 0.2, "sources": [ "GDPval-AA · Artificial Analysis", "AA-Briefcase · Artificial Analysis", "AutomationBench-AA · Artificial Analysis", "Finance Agent · Vals AI", "APEX-Agents · Mercor", "Harvey 法律 Agent 基准 · Vals AI", "τ³-Banking · Mercor", "LiveBench Data Analysis · LiveBench", "MedCode · Vals AI", "AA-Analyst Agent · Artificial Analysis" ] }, { "key": "knowledge", "name": "知识问答", "weight": 0.15, "sources": [ "AA-Omniscience · Artificial Analysis", "SimpleQA Verified · Epoch AI", "AA-LCR · Artificial Analysis", "MLCR-AA · Artificial Analysis", "LiveBench Language · LiveBench", "LiveBench IF · LiveBench", "BullshitBench v2 · BullshitBench" ] }, { "key": "preference", "name": "真人盲选", "weight": 0.05, "sources": [ "Arena Text · LMArena" ] } ], "anchors": [], "domains": [ { "key": "coding", "name": "编程", "weight": 0.4, "benchmarks": [ { "name": "Terminal-Bench 4.0", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-terminal-bench", "suspended": null }, { "name": "Terminal-Bench 4.0", "operator": "Vals AI", "sourceKey": "vals-terminal-bench-4", "suspended": null }, { "name": "LiveBench Coding", "operator": "LiveBench", "sourceKey": "livebench-coding", "suspended": null }, { "name": "SciCode", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-scicode", "suspended": null }, { "name": "Code Migration", "operator": "Vals AI", "sourceKey": "vals-migration", "suspended": null }, { "name": "ProgramBench", "operator": "Vals AI", "sourceKey": "vals-program", "suspended": null }, { "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "sourceKey": "vals-vibe-code", "suspended": null }, { "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "sourceKey": "vals-vcb-1-100", "suspended": null }, { "name": "CyberBench Patch v1.1", "operator": "Vals AI", "sourceKey": "vals-cyber-patch", "suspended": null } ] }, { "key": "reasoning", "name": "科研", "weight": 0.2, "benchmarks": [ { "name": "Humanity's Last Exam", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-hle", "suspended": null }, { "name": "FrontierMath T1–3", "operator": "Epoch AI", "sourceKey": "epoch-frontiermath", "suspended": null }, { "name": "LiveBench Reasoning", "operator": "LiveBench", "sourceKey": "livebench-reasoning", "suspended": null }, { "name": "CritPt", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-critpt", "suspended": null }, { "name": "Chess Puzzles", "operator": "Epoch AI", "sourceKey": "epoch-chess", "suspended": null }, { "name": "Mystery Game Puzzles", "operator": "Epoch AI", "sourceKey": "epoch-mystery", "suspended": null }, { "name": "MysteryMechanism", "operator": "Vals AI", "sourceKey": "vals-mysterymechanism", "suspended": null }, { "name": "Terminal-Bench Science", "operator": "Vals AI", "sourceKey": "vals-terminal-bench-science", "suspended": null }, { "name": "Terminal-Bench Science", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-terminal-bench-science", "suspended": null } ] }, { "key": "professional", "name": "专业办公", "weight": 0.2, "benchmarks": [ { "name": "GDPval-AA", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-gdpval", "suspended": null }, { "name": "AA-Briefcase", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-briefcase", "suspended": null }, { "name": "AutomationBench-AA", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-automationbench", "suspended": null }, { "name": "Finance Agent", "operator": "Vals AI", "sourceKey": "vals-finance-agent", "suspended": null }, { "name": "APEX-Agents", "operator": "Mercor", "sourceKey": "mercor-apex-agents", "suspended": null }, { "name": "Harvey 法律 Agent 基准", "operator": "Vals AI", "sourceKey": "vals-harvey-lab", "suspended": null }, { "name": "τ³-Banking", "operator": "Mercor", "sourceKey": "mercor-tau3-banking", "suspended": null }, { "name": "LiveBench Data Analysis", "operator": "LiveBench", "sourceKey": "livebench-data-analysis", "suspended": null }, { "name": "MedCode", "operator": "Vals AI", "sourceKey": "vals-medcode", "suspended": null }, { "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-analyst-agent", "suspended": null } ] }, { "key": "knowledge", "name": "知识问答", "weight": 0.15, "benchmarks": [ { "name": "AA-Omniscience", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-omniscience", "suspended": null }, { "name": "SimpleQA Verified", "operator": "Epoch AI", "sourceKey": "epoch-simpleqa", "suspended": null }, { "name": "AA-LCR", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-lcr", "suspended": null }, { "name": "MLCR-AA", "operator": "Artificial Analysis", "sourceKey": "artificial-analysis-mlcr", "suspended": null }, { "name": "LiveBench Language", "operator": "LiveBench", "sourceKey": "livebench-writing", "suspended": null }, { "name": "LiveBench IF", "operator": "LiveBench", "sourceKey": "livebench-writing", "suspended": null }, { "name": "BullshitBench v2", "operator": "BullshitBench", "sourceKey": "bullshitbench-v2", "suspended": null } ] }, { "key": "preference", "name": "真人盲选", "weight": 0.05, "benchmarks": [ { "name": "Arena Text", "operator": "LMArena", "sourceKey": "arena-text", "suspended": null } ] } ], "reference_model_count": 92, "score_scale": { "mean": 50, "points_per_sd": 15, "reference": "frozen_reference_group" }, "fx": { "rate": 6.7023, "asOf": "2026-10-08", "sourceName": "欧洲央行" } }, "mode": "aihot", "default_metric": "professional", "as_of": "2026-10-09T00:05:49.189Z", "fetched_at": "2026-10-09T04:17:12.573Z", "run_id": "yiu8g6ysf3xdoq8xed0giima0", "methodology_version": "2026.10-capability-scale-v19", "metrics": [ { "id": "overall", "label": "综合", "source_count": 35, "operator_count": 7, "model_count": 30, "model_count_basis": "returned_ranked_entries", "url": "https://aihot.news/leaderboard" }, { "id": "coding", "label": "编程", "source_count": 9, "operator_count": 3, "model_count": 30, "model_count_basis": "returned_ranked_entries", "url": "https://aihot.news/leaderboard/category/coding" }, { "id": "reasoning", "label": "科研", "source_count": 9, "operator_count": 4, "model_count": 30, "model_count_basis": "returned_ranked_entries", "url": "https://aihot.news/leaderboard/category/reasoning" }, { "id": "knowledge", "label": "知识问答", "source_count": 6, "operator_count": 4, "model_count": 30, "model_count_basis": "returned_ranked_entries", "url": "https://aihot.news/leaderboard/category/knowledge" }, { "id": "professional", "label": "专业办公", "source_count": 10, "operator_count": 4, "model_count": 30, "model_count_basis": "returned_ranked_entries", "url": "https://aihot.news/leaderboard/category/professional" } ], "boards": { "overall": [ { "id": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "rank": 1, "score": 73.9, "source_count": 33, "confidence": null, "coverage": 0.964, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/models/opus-5-5/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5-5" }, { "id": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "rank": 2, "score": 73.5, "source_count": 21, "confidence": null, "coverage": 0.679, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-4-argon" }, { "id": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "rank": 3, "score": 72.3, "source_count": 35, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 1.675575, "official_url": "https://platform.claude.com/docs/en/models/fable-5-1/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5-1" }, { "id": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "rank": 4, "score": 71.3, "source_count": 35, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-astra" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-astra" }, { "id": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "rank": 5, "score": 71.3, "source_count": 33, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5-5" }, { "id": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "rank": 6, "score": 69.9, "source_count": 30, "confidence": null, "coverage": 0.893, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5" }, { "id": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "rank": 7, "score": 69, "source_count": 34, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 0.6702300000000001, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-1-sol" }, { "id": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "rank": 8, "score": 68.9, "source_count": 35, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5" }, { "id": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "rank": 9, "score": 65.9, "source_count": 33, "confidence": null, "coverage": 0.929, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 2.6809200000000004, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-sol" }, { "id": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "rank": 10, "score": 65.7, "source_count": 32, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-sol" }, { "id": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "rank": 11, "score": 65.4, "source_count": 30, "confidence": null, "coverage": 0.893, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-3" }, { "id": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "rank": 12, "score": 64.5, "source_count": 33, "confidence": null, "coverage": 0.964, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-8-flash" }, { "id": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "rank": 13, "score": 63.3, "source_count": 32, "confidence": null, "coverage": 0.929, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "input_cny": 20, "output_cny": 100, "cached_cny": 2, "official_url": "https://platform.moonshot.cn/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/moonshotai/Kimi-K3" }, "url": "https://aihot.news/leaderboard/kimi-k-3" }, { "id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "rank": 14, "score": 63, "source_count": 30, "confidence": null, "coverage": 0.893, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-8" }, { "id": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "rank": 15, "score": 62.9, "source_count": 23, "confidence": null, "coverage": 0.679, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "input_cny": 3, "output_cny": 6, "cached_cny": 0.025, "official_url": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-pro" }, { "id": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "rank": 16, "score": 62.8, "source_count": 31, "confidence": null, "coverage": 0.929, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-7-flash" }, { "id": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "rank": 17, "score": 62.4, "source_count": 34, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "input_cny": 9.38322, "output_cny": 29.490120000000005, "cached_cny": 1.742598, "official_url": "https://docs.z.ai/guides/overview/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3" }, "url": "https://aihot.news/leaderboard/glm-5-3" }, { "id": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "rank": 18, "score": 62.3, "source_count": 28, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "input_cny": 33.5115, "output_cny": 201.06900000000002, "cached_cny": 3.35115, "official_url": "https://developers.openai.com/api/docs/models/gpt-5.5" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-5" }, { "id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "rank": 19, "score": 62.1, "source_count": 19, "confidence": null, "coverage": 0.643, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-7" }, { "id": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "rank": 20, "score": 62.1, "source_count": 33, "confidence": null, "coverage": 0.964, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-terra" }, { "id": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "rank": 21, "score": 61.6, "source_count": 31, "confidence": null, "coverage": 0.964, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-7" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-7" }, { "id": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "rank": 22, "score": 61.3, "source_count": 35, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-6" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-6" }, { "id": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "rank": 23, "score": 61.1, "source_count": 26, "confidence": null, "coverage": 0.821, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-2" }, { "id": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "rank": 24, "score": 60.6, "source_count": 20, "confidence": null, "coverage": 0.571, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/step-5-preview" }, { "id": "claude-haiku-5-5", "name": "Claude Haiku 5.5", "provider": "Anthropic", "rank": 25, "score": 60.6, "source_count": 27, "confidence": null, "coverage": 0.821, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-haiku-5-5" }, { "id": "deepseek-v-4-1-flash", "name": "DeepSeek V4.1 Flash", "provider": "DeepSeek", "rank": 26, "score": 60.1, "source_count": 30, "confidence": null, "coverage": 0.929, "price": { "currency": "USD", "input": 0.3, "output": 1.2, "cached": 0.006, "input_cny": 2.01069, "output_cny": 8.04276, "cached_cny": 0.0402138, "official_url": "https://api-docs.deepseek.com/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-1-flash" }, { "id": "deepseek-v-4-pro-20260813", "name": "DeepSeek V4 Pro 0813", "provider": "DeepSeek", "rank": 27, "score": 58.8, "source_count": 33, "confidence": null, "coverage": 0.964, "price": { "currency": "CNY", "input": 9, "output": 27, "cached": 0.3, "input_cny": 9, "output_cny": 27, "cached_cny": 0.3, "official_url": "https://api-docs.deepseek.com/zh-cn/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-pro-20260813" }, { "id": "claude-sonnet-5", "name": "Claude Sonnet 5", "provider": "Anthropic", "rank": 28, "score": 58.8, "source_count": 32, "confidence": null, "coverage": 0.929, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5" }, { "id": "qwen-3-8-max", "name": "Qwen3.8 Max", "provider": "Alibaba", "rank": 29, "score": 58.6, "source_count": 33, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 1.5, "input_cny": 12, "output_cny": 36, "cached_cny": 1.5, "official_url": "https://help.aliyun.com/zh/model-studio/qwen3-8-max" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/qwen-3-8-max" }, { "id": "muse-spark-1-1", "name": "Muse Spark 1.1", "provider": "Meta", "rank": 30, "score": 58.5, "source_count": 21, "confidence": null, "coverage": 0.714, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-1" } ], "coding": [ { "id": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "rank": 1, "score": 74.3, "source_count": 8, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/models/opus-5-5/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5-5?from=coding" }, { "id": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "rank": 2, "score": 73.9, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5-5?from=coding" }, { "id": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "rank": 3, "score": 71.7, "source_count": 7, "confidence": null, "coverage": 0.857, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-4-argon?from=coding" }, { "id": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "rank": 4, "score": 71.7, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 1.675575, "official_url": "https://platform.claude.com/docs/en/models/fable-5-1/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5-1?from=coding" }, { "id": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "rank": 5, "score": 69.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-astra" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-astra?from=coding" }, { "id": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "rank": 6, "score": 69.2, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5?from=coding" }, { "id": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "rank": 7, "score": 67, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5?from=coding" }, { "id": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "rank": 8, "score": 66.5, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-sol?from=coding" }, { "id": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "rank": 9, "score": 66.4, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 0.6702300000000001, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-1-sol?from=coding" }, { "id": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "rank": 10, "score": 65.8, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 2.6809200000000004, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-sol?from=coding" }, { "id": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "rank": 11, "score": 64.1, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "input_cny": 3, "output_cny": 6, "cached_cny": 0.025, "official_url": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-pro?from=coding" }, { "id": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "rank": 12, "score": 63.4, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "input_cny": 9.38322, "output_cny": 29.490120000000005, "cached_cny": 1.742598, "official_url": "https://docs.z.ai/guides/overview/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3" }, "url": "https://aihot.news/leaderboard/glm-5-3?from=coding" }, { "id": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "rank": 13, "score": 63.3, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-3?from=coding" }, { "id": "claude-haiku-5-5", "name": "Claude Haiku 5.5", "provider": "Anthropic", "rank": 14, "score": 62.7, "source_count": 9, "confidence": null, "coverage": 1, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-haiku-5-5?from=coding" }, { "id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "rank": 15, "score": 62.3, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-8?from=coding" }, { "id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "rank": 16, "score": 62.3, "source_count": 4, "confidence": null, "coverage": 0.571, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-7?from=coding" }, { "id": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "rank": 17, "score": 61.7, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-7" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-7?from=coding" }, { "id": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "rank": 18, "score": 61.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-terra?from=coding" }, { "id": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "rank": 19, "score": 61.4, "source_count": 5, "confidence": null, "coverage": 0.571, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/step-5-preview?from=coding" }, { "id": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "rank": 20, "score": 61.3, "source_count": 8, "confidence": null, "coverage": 0.857, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "input_cny": 20, "output_cny": 100, "cached_cny": 2, "official_url": "https://platform.moonshot.cn/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/moonshotai/Kimi-K3" }, "url": "https://aihot.news/leaderboard/kimi-k-3?from=coding" }, { "id": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "rank": 21, "score": 60.9, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-8-flash?from=coding" }, { "id": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "rank": 22, "score": 60.9, "source_count": 6, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "input_cny": 33.5115, "output_cny": 201.06900000000002, "cached_cny": 3.35115, "official_url": "https://developers.openai.com/api/docs/models/gpt-5.5" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-5?from=coding" }, { "id": "deepseek-v-4-1-flash", "name": "DeepSeek V4.1 Flash", "provider": "DeepSeek", "rank": 23, "score": 60.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.3, "output": 1.2, "cached": 0.006, "input_cny": 2.01069, "output_cny": 8.04276, "cached_cny": 0.0402138, "official_url": "https://api-docs.deepseek.com/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-1-flash?from=coding" }, { "id": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "rank": 24, "score": 60.3, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-7-flash?from=coding" }, { "id": "gpt-6-luna", "name": "GPT-6 Luna", "provider": "OpenAI", "rank": 25, "score": 59.6, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.1, "output": 0.5, "cached": 0.01, "input_cny": 0.6702300000000001, "output_cny": 3.35115, "cached_cny": 0.067023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-luna" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-luna?from=coding" }, { "id": "mimo-v-2-6-flash", "name": "MiMo V 2.6 Flash", "provider": "Xiaomi", "rank": 26, "score": 59.3, "source_count": 7, "confidence": null, "coverage": 0.857, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-flash?from=coding" }, { "id": "gpt-5-6-luna", "name": "GPT-5.6 Luna", "provider": "OpenAI", "rank": 27, "score": 59.2, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.2, "output": 1.2, "cached": 0.02, "input_cny": 1.3404600000000002, "output_cny": 8.04276, "cached_cny": 0.134046, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-luna?from=coding" }, { "id": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "rank": 28, "score": 58.9, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-2?from=coding" }, { "id": "deepseek-v-4-pro-20260813", "name": "DeepSeek V4 Pro 0813", "provider": "DeepSeek", "rank": 29, "score": 58.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 9, "output": 27, "cached": 0.3, "input_cny": 9, "output_cny": 27, "cached_cny": 0.3, "official_url": "https://api-docs.deepseek.com/zh-cn/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-pro-20260813?from=coding" }, { "id": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "rank": 30, "score": 58.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-6" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-6?from=coding" } ], "reasoning": [ { "id": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "rank": 1, "score": 77.6, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-astra" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-astra?from=reasoning" }, { "id": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "rank": 2, "score": 76.8, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/models/opus-5-5/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5-5?from=reasoning" }, { "id": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "rank": 3, "score": 75.7, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 0.6702300000000001, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-1-sol?from=reasoning" }, { "id": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "rank": 4, "score": 74.7, "source_count": 4, "confidence": null, "coverage": 0.571, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-4-argon?from=reasoning" }, { "id": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "rank": 5, "score": 74.3, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5-5?from=reasoning" }, { "id": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "rank": 6, "score": 74.2, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 1.675575, "official_url": "https://platform.claude.com/docs/en/models/fable-5-1/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5-1?from=reasoning" }, { "id": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "rank": 7, "score": 70.8, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5?from=reasoning" }, { "id": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "rank": 8, "score": 70.1, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 2.6809200000000004, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-sol?from=reasoning" }, { "id": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "rank": 9, "score": 69.8, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5?from=reasoning" }, { "id": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "rank": 10, "score": 69.3, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-sol?from=reasoning" }, { "id": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "rank": 11, "score": 66.1, "source_count": 8, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-terra?from=reasoning" }, { "id": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "rank": 12, "score": 66, "source_count": 6, "confidence": null, "coverage": 0.714, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "input_cny": 33.5115, "output_cny": 201.06900000000002, "cached_cny": 3.35115, "official_url": "https://developers.openai.com/api/docs/models/gpt-5.5" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-5?from=reasoning" }, { "id": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "rank": 13, "score": 65.3, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-8-flash?from=reasoning" }, { "id": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "rank": 14, "score": 65.1, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-3?from=reasoning" }, { "id": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "rank": 15, "score": 63.2, "source_count": 3, "confidence": null, "coverage": 0.429, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-2?from=reasoning" }, { "id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "rank": 16, "score": 63.1, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-8?from=reasoning" }, { "id": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "rank": 17, "score": 62.8, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "input_cny": 20, "output_cny": 100, "cached_cny": 2, "official_url": "https://platform.moonshot.cn/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/moonshotai/Kimi-K3" }, "url": "https://aihot.news/leaderboard/kimi-k-3?from=reasoning" }, { "id": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "rank": 18, "score": 62.7, "source_count": 7, "confidence": null, "coverage": 0.857, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-7-flash?from=reasoning" }, { "id": "claude-haiku-5-5", "name": "Claude Haiku 5.5", "provider": "Anthropic", "rank": 19, "score": 62.6, "source_count": 6, "confidence": null, "coverage": 0.714, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-haiku-5-5?from=reasoning" }, { "id": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "rank": 20, "score": 62.3, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-6" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-6?from=reasoning" }, { "id": "qwen-3-8-max", "name": "Qwen3.8 Max", "provider": "Alibaba", "rank": 21, "score": 62.2, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 1.5, "input_cny": 12, "output_cny": 36, "cached_cny": 1.5, "official_url": "https://help.aliyun.com/zh/model-studio/qwen3-8-max" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/qwen-3-8-max?from=reasoning" }, { "id": "muse-spark-1-1", "name": "Muse Spark 1.1", "provider": "Meta", "rank": 22, "score": 60.9, "source_count": 3, "confidence": null, "coverage": 0.429, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-1?from=reasoning" }, { "id": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "rank": 23, "score": 60.9, "source_count": 5, "confidence": null, "coverage": 0.571, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "input_cny": 3, "output_cny": 6, "cached_cny": 0.025, "official_url": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-pro?from=reasoning" }, { "id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "rank": 24, "score": 60.5, "source_count": 4, "confidence": null, "coverage": 0.429, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-7?from=reasoning" }, { "id": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "rank": 25, "score": 60.3, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "input_cny": 9.38322, "output_cny": 29.490120000000005, "cached_cny": 1.742598, "official_url": "https://docs.z.ai/guides/overview/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3" }, "url": "https://aihot.news/leaderboard/glm-5-3?from=reasoning" }, { "id": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "rank": 26, "score": 60, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-7" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-7?from=reasoning" }, { "id": "deepseek-v-4-pro-20260813", "name": "DeepSeek V4 Pro 0813", "provider": "DeepSeek", "rank": 27, "score": 59.8, "source_count": 8, "confidence": null, "coverage": 0.857, "price": { "currency": "CNY", "input": 9, "output": 27, "cached": 0.3, "input_cny": 9, "output_cny": 27, "cached_cny": 0.3, "official_url": "https://api-docs.deepseek.com/zh-cn/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-pro-20260813?from=reasoning" }, { "id": "grok-4-5", "name": "Grok 4.5", "provider": "xAI", "rank": 28, "score": 59.6, "source_count": 5, "confidence": null, "coverage": 0.714, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.3, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 2.01069, "official_url": "https://docs.x.ai/developers/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-5?from=reasoning" }, { "id": "gemini-3-1-pro-preview", "name": "Gemini 3.1 Pro Preview", "provider": "Google", "rank": 29, "score": 59.3, "source_count": 8, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-1-pro-preview?from=reasoning" }, { "id": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "rank": 30, "score": 59.2, "source_count": 5, "confidence": null, "coverage": 0.571, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/step-5-preview?from=reasoning" } ], "knowledge": [ { "id": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "rank": 1, "score": 74.8, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-astra" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-astra?from=knowledge" }, { "id": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "rank": 2, "score": 74.5, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 1.675575, "official_url": "https://platform.claude.com/docs/en/models/fable-5-1/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5-1?from=knowledge" }, { "id": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "rank": 3, "score": 73.6, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5?from=knowledge" }, { "id": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "rank": 4, "score": 72.7, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/models/opus-5-5/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5-5?from=knowledge" }, { "id": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "rank": 5, "score": 72.4, "source_count": 2, "confidence": null, "coverage": 0.4, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-4-argon?from=knowledge" }, { "id": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "rank": 6, "score": 71.9, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5?from=knowledge" }, { "id": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "rank": 7, "score": 71.8, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 0.6702300000000001, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-1-sol?from=knowledge" }, { "id": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "rank": 8, "score": 71.1, "source_count": 5, "confidence": null, "coverage": 0.8, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-8-flash?from=knowledge" }, { "id": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "rank": 9, "score": 68.1, "source_count": 4, "confidence": null, "coverage": 0.8, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-3?from=knowledge" }, { "id": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "rank": 10, "score": 68, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "input_cny": 20, "output_cny": 100, "cached_cny": 2, "official_url": "https://platform.moonshot.cn/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/moonshotai/Kimi-K3" }, "url": "https://aihot.news/leaderboard/kimi-k-3?from=knowledge" }, { "id": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "rank": 11, "score": 65.6, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 2.6809200000000004, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-sol?from=knowledge" }, { "id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "rank": 12, "score": 65.5, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-8?from=knowledge" }, { "id": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "rank": 13, "score": 65.4, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-7-flash?from=knowledge" }, { "id": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "rank": 14, "score": 64, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-sol?from=knowledge" }, { "id": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "rank": 15, "score": 63.4, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5-5?from=knowledge" }, { "id": "gemini-3-1-pro-preview", "name": "Gemini 3.1 Pro Preview", "provider": "Google", "rank": 16, "score": 63.3, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-1-pro-preview?from=knowledge" }, { "id": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "rank": 17, "score": 63.3, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "input_cny": 33.5115, "output_cny": 201.06900000000002, "cached_cny": 3.35115, "official_url": "https://developers.openai.com/api/docs/models/gpt-5.5" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-5?from=knowledge" }, { "id": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "rank": 18, "score": 62.5, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "input_cny": 9.38322, "output_cny": 29.490120000000005, "cached_cny": 1.742598, "official_url": "https://docs.z.ai/guides/overview/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3" }, "url": "https://aihot.news/leaderboard/glm-5-3?from=knowledge" }, { "id": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "rank": 19, "score": 62.5, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-6" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-6?from=knowledge" }, { "id": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "rank": 20, "score": 62.3, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-2?from=knowledge" }, { "id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "rank": 21, "score": 62.2, "source_count": 4, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-7?from=knowledge" }, { "id": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "rank": 22, "score": 62.1, "source_count": 5, "confidence": null, "coverage": 0.8, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-7" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-7?from=knowledge" }, { "id": "gemini-3-6-flash", "name": "Gemini 3.6 Flash", "provider": "Google", "rank": 23, "score": 61, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-6-flash?from=knowledge" }, { "id": "grok-4-5", "name": "Grok 4.5", "provider": "xAI", "rank": 24, "score": 60.7, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.3, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 2.01069, "official_url": "https://docs.x.ai/developers/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-5?from=knowledge" }, { "id": "gemini-3-5-flash", "name": "Gemini 3.5 Flash", "provider": "Google", "rank": 25, "score": 60.6, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.5, "output": 9, "cached": 0.15, "input_cny": 10.05345, "output_cny": 60.3207, "cached_cny": 1.005345, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-5-flash?from=knowledge" }, { "id": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "rank": 26, "score": 60.2, "source_count": 3, "confidence": null, "coverage": 0.4, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/step-5-preview?from=knowledge" }, { "id": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "rank": 27, "score": 59.8, "source_count": 6, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-terra?from=knowledge" }, { "id": "deepseek-v-4-1-flash", "name": "DeepSeek V4.1 Flash", "provider": "DeepSeek", "rank": 28, "score": 59.8, "source_count": 5, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.3, "output": 1.2, "cached": 0.006, "input_cny": 2.01069, "output_cny": 8.04276, "cached_cny": 0.0402138, "official_url": "https://api-docs.deepseek.com/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-1-flash?from=knowledge" }, { "id": "qwen-3-7-max", "name": "Qwen3.7 Max", "provider": "Alibaba", "rank": 29, "score": 59.7, "source_count": 5, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 2.4, "input_cny": 12, "output_cny": 36, "cached_cny": 2.4, "official_url": "https://help.aliyun.com/zh/model-studio/model-pricing" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/qwen-3-7-max?from=knowledge" }, { "id": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "rank": 30, "score": 59.7, "source_count": 3, "confidence": null, "coverage": 0.4, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "input_cny": 3, "output_cny": 6, "cached_cny": 0.025, "official_url": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-pro?from=knowledge" } ], "professional": [ { "id": "gemini-4-argon", "name": "Gemini 4 Argon", "provider": "Google", "rank": 1, "score": 76.4, "source_count": 7, "confidence": null, "coverage": 0.75, "price": null, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-4-argon?from=professional" }, { "id": "claude-opus-5-5", "name": "Claude Opus 5.5", "provider": "Anthropic", "rank": 2, "score": 71.3, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.2, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/models/opus-5-5/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5-5?from=professional" }, { "id": "claude-sonnet-5-5", "name": "Claude Sonnet 5.5", "provider": "Anthropic", "rank": 3, "score": 70.6, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-sonnet-5-5?from=professional" }, { "id": "claude-fable-5-1", "name": "Claude Fable 5.1", "provider": "Anthropic", "rank": 4, "score": 70.2, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 0.25, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 1.675575, "official_url": "https://platform.claude.com/docs/en/models/fable-5-1/overview" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5-1?from=professional" }, { "id": "claude-opus-5", "name": "Claude Opus 5", "provider": "Anthropic", "rank": 5, "score": 69, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-5?from=professional" }, { "id": "claude-fable-5", "name": "Claude Fable 5", "provider": "Anthropic", "rank": 6, "score": 68.7, "source_count": 9, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-fable-5?from=professional" }, { "id": "muse-spark-1-3", "name": "Muse Spark 1.3", "provider": "Meta", "rank": 7, "score": 67.5, "source_count": 7, "confidence": null, "coverage": 0.75, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-3?from=professional" }, { "id": "gpt-6-astra", "name": "GPT-6 Astra", "provider": "OpenAI", "rank": 8, "score": 67.2, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 10, "output": 50, "cached": 1, "input_cny": 67.023, "output_cny": 335.115, "cached_cny": 6.7023, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-astra" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-astra?from=professional" }, { "id": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "provider": "Google", "rank": 9, "score": 65.9, "source_count": 9, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-7-flash?from=professional" }, { "id": "grok-4-6", "name": "Grok 4.6", "provider": "xAI", "rank": 10, "score": 65.8, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-6" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-6?from=professional" }, { "id": "gpt-6-1-sol", "name": "GPT-6.1 Sol", "provider": "OpenAI", "rank": 11, "score": 65.8, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.1, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 0.6702300000000001, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-1-sol?from=professional" }, { "id": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "provider": "Google", "rank": 12, "score": 65.3, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.75, "output": 3.75, "cached": 0.075, "input_cny": 5.026725, "output_cny": 25.133625000000002, "cached_cny": 0.5026725, "official_url": "https://ai.google.dev/gemini-api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gemini-3-8-flash?from=professional" }, { "id": "mimo-v-2-6-pro", "name": "MiMo-V2.6-Pro", "provider": "Xiaomi", "rank": 13, "score": 65.2, "source_count": 7, "confidence": null, "coverage": 0.75, "price": { "currency": "CNY", "input": 3, "output": 6, "cached": 0.025, "input_cny": 3, "output_cny": 6, "cached_cny": 0.025, "official_url": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-pro?from=professional" }, { "id": "grok-4-7", "name": "Grok 4.7", "provider": "xAI", "rank": 14, "score": 64.4, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.5, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 3.35115, "official_url": "https://docs.x.ai/developers/grok-4-7" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-7?from=professional" }, { "id": "kimi-k-3", "name": "Kimi K3", "provider": "Moonshot AI", "rank": 15, "score": 63.9, "source_count": 10, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 20, "output": 100, "cached": 2, "input_cny": 20, "output_cny": 100, "cached_cny": 2, "official_url": "https://platform.moonshot.cn/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/moonshotai/Kimi-K3" }, "url": "https://aihot.news/leaderboard/kimi-k-3?from=professional" }, { "id": "qwen-3-8-max", "name": "Qwen3.8 Max", "provider": "Alibaba", "rank": 16, "score": 63.5, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "CNY", "input": 12, "output": 36, "cached": 1.5, "input_cny": 12, "output_cny": 36, "cached_cny": 1.5, "official_url": "https://help.aliyun.com/zh/model-studio/qwen3-8-max" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/qwen-3-8-max?from=professional" }, { "id": "gpt-6-sol", "name": "GPT-6 Sol", "provider": "OpenAI", "rank": 17, "score": 62.9, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 10, "cached": 0.2, "input_cny": 13.4046, "output_cny": 67.023, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models/gpt-6-sol" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-6-sol?from=professional" }, { "id": "mimo-v-2-6-flash", "name": "MiMo V 2.6 Flash", "provider": "Xiaomi", "rank": 18, "score": 62.8, "source_count": 6, "confidence": null, "coverage": 0.625, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/mimo-v-2-6-flash?from=professional" }, { "id": "glm-5-3", "name": "GLM-5.3", "provider": "Z.ai", "rank": 19, "score": 62.8, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 1.4, "output": 4.4, "cached": 0.26, "input_cny": 9.38322, "output_cny": 29.490120000000005, "cached_cny": 1.742598, "official_url": "https://docs.z.ai/guides/overview/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3" }, "url": "https://aihot.news/leaderboard/glm-5-3?from=professional" }, { "id": "glm-5-3-flash", "name": "GLM-5.3 Flash", "provider": "Z.ai", "rank": 20, "score": 62.7, "source_count": 8, "confidence": null, "coverage": 0.875, "price": { "currency": "CNY", "input": 0.8, "output": 2.8, "cached": 0.23, "input_cny": 0.8, "output_cny": 2.8, "cached_cny": 0.23, "official_url": "https://www.bigmodel.cn/pricing" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/zai-org/GLM-5.3-Flash" }, "url": "https://aihot.news/leaderboard/glm-5-3-flash?from=professional" }, { "id": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "provider": "OpenAI", "rank": 21, "score": 62.7, "source_count": 8, "confidence": null, "coverage": 0.75, "price": { "currency": "USD", "input": 4, "output": 20, "cached": 0.4, "input_cny": 26.8092, "output_cny": 134.046, "cached_cny": 2.6809200000000004, "official_url": "https://developers.openai.com/api/docs/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-sol?from=professional" }, { "id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "Anthropic", "rank": 22, "score": 62.6, "source_count": 6, "confidence": null, "coverage": 0.625, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-7?from=professional" }, { "id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "Anthropic", "rank": 23, "score": 62.4, "source_count": 9, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 5, "output": 25, "cached": 0.5, "input_cny": 33.5115, "output_cny": 167.5575, "cached_cny": 3.35115, "official_url": "https://platform.claude.com/docs/en/about-claude/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/claude-opus-4-8?from=professional" }, { "id": "qwen-3-8-flash-next", "name": "Qwen3.8 Flash Next", "provider": "Alibaba", "rank": 24, "score": 62.2, "source_count": 4, "confidence": null, "coverage": 0.375, "price": { "currency": "CNY", "input": 0.8, "output": 2.7, "cached": 0.1, "input_cny": 0.8, "output_cny": 2.7, "cached_cny": 0.1, "official_url": "https://help.aliyun.com/zh/model-studio/qwen3-8-flash" }, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/qwen-3-8-flash-next?from=professional" }, { "id": "muse-spark-1-2", "name": "Muse Spark 1.2", "provider": "Meta", "rank": 25, "score": 62.1, "source_count": 8, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 1.25, "output": 4.25, "cached": 0.15, "input_cny": 8.377875, "output_cny": 28.484775, "cached_cny": 1.005345, "official_url": "https://dev.meta.ai/docs/pricing-rate-limits/" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/muse-spark-1-2?from=professional" }, { "id": "step-5-preview", "name": "Step 5 Preview", "provider": "StepFun", "rank": 26, "score": 61.8, "source_count": 6, "confidence": null, "coverage": 0.625, "price": null, "access": { "domestic": true, "weights_url": null }, "url": "https://aihot.news/leaderboard/step-5-preview?from=professional" }, { "id": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "provider": "OpenAI", "rank": 27, "score": 61.6, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 2, "output": 12, "cached": 0.2, "input_cny": 13.4046, "output_cny": 80.4276, "cached_cny": 1.3404600000000002, "official_url": "https://developers.openai.com/api/docs/models" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-6-terra?from=professional" }, { "id": "deepseek-v-4-1-flash", "name": "DeepSeek V4.1 Flash", "provider": "DeepSeek", "rank": 28, "score": 61.1, "source_count": 9, "confidence": null, "coverage": 1, "price": { "currency": "USD", "input": 0.3, "output": 1.2, "cached": 0.006, "input_cny": 2.01069, "output_cny": 8.04276, "cached_cny": 0.0402138, "official_url": "https://api-docs.deepseek.com/quick_start/pricing/" }, "access": { "domestic": true, "weights_url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash" }, "url": "https://aihot.news/leaderboard/deepseek-v-4-1-flash?from=professional" }, { "id": "gpt-5-5", "name": "GPT-5.5", "provider": "OpenAI", "rank": 29, "score": 60.9, "source_count": 9, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 5, "output": 30, "cached": 0.5, "input_cny": 33.5115, "output_cny": 201.06900000000002, "cached_cny": 3.35115, "official_url": "https://developers.openai.com/api/docs/models/gpt-5.5" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/gpt-5-5?from=professional" }, { "id": "grok-4-5", "name": "Grok 4.5", "provider": "xAI", "rank": 30, "score": 60.7, "source_count": 9, "confidence": null, "coverage": 0.875, "price": { "currency": "USD", "input": 2, "output": 6, "cached": 0.3, "input_cny": 13.4046, "output_cny": 40.2138, "cached_cny": 2.01069, "official_url": "https://docs.x.ai/developers/pricing" }, "access": { "domestic": false, "weights_url": null }, "url": "https://aihot.news/leaderboard/grok-4-5?from=professional" } ] }, "pending": { "overall": [ { "id": "qwen-3-8-max-0902", "name": "Qwen3.8 Max 0902", "provider": "Alibaba", "reason": "编程评测还不到 3 类" }, { "id": "ember-1", "name": "ember-1", "provider": "Fireworks AI", "reason": "目前只有一家评测机构的成绩" }, { "id": "ling-3-1-flash", "name": "Ling 3.1 Flash", "provider": "InclusionAI", "reason": "目前只有一家评测机构的成绩" }, { "id": "gpt-5-4", "name": "GPT-5.4", "provider": "OpenAI", "reason": "编程评测还不到 3 类" }, { "id": "claude-opus-4-6", "name": "Claude Opus 4.6", "provider": "Anthropic", "reason": "编程评测还不到 3 类" } ], "coding": [ { "id": "ember-1", "name": "ember-1", "provider": "Fireworks AI", "reason": "目前只有一家评测机构的成绩" } ], "reasoning": [ { "id": "gpt-5-4-2026-03-05", "name": "GPT-5.4 2026-03-05", "provider": "OpenAI", "reason": "知识问答评测还不到 2 类" }, { "id": "qwen-3-8-max-0902", "name": "Qwen3.8 Max 0902", "provider": "Alibaba", "reason": "编程评测还不到 3 类" }, { "id": "qwen-3-8-2-4-t-a-95-b", "name": "Qwen3.8 2.4 T A 95 B", "provider": "Alibaba", "reason": "目前只有一家评测机构的成绩" }, { "id": "ling-3-1-flash", "name": "Ling 3.1 Flash", "provider": "InclusionAI", "reason": "目前只有一家评测机构的成绩" }, { "id": "ember-1", "name": "ember-1", "provider": "Fireworks AI", "reason": "目前只有一家评测机构的成绩" }, { "id": "claude-opus-4-6", "name": "Claude Opus 4.6", "provider": "Anthropic", "reason": "编程评测还不到 3 类" } ], "knowledge": [ { "id": "claude-opus-4-6", "name": "Claude Opus 4.6", "provider": "Anthropic", "reason": "编程评测还不到 3 类" } ], "professional": [ { "id": "qwen-3-8-max-0902", "name": "Qwen3.8 Max 0902", "provider": "Alibaba", "reason": "编程评测还不到 3 类" }, { "id": "ling-3-1-flash", "name": "Ling 3.1 Flash", "provider": "InclusionAI", "reason": "目前只有一家评测机构的成绩" }, { "id": "gpt-5-4", "name": "GPT-5.4", "provider": "OpenAI", "reason": "编程评测还不到 3 类" }, { "id": "ember-1", "name": "ember-1", "provider": "Fireworks AI", "reason": "目前只有一家评测机构的成绩" }, { "id": "qwen-3-8-2-4-t-a-95-b", "name": "Qwen3.8 2.4 T A 95 B", "provider": "Alibaba", "reason": "目前只有一家评测机构的成绩" }, { "id": "mistral-large-4-preview", "name": "Mistral Large 4 Preview", "provider": "Mistral AI", "reason": "目前只有一家评测机构的成绩" } ] }, "sources": [ { "key": "artificial-analysis-terminal-bench", "name": "Terminal-Bench 4.0 · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/terminalbench-4-0", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "在命令行环境里独立完成一整件工程任务。" }, { "key": "vals-terminal-bench-4", "name": "Terminal-Bench 4.0 · Vals", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/terminal-bench-4", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "Vals 用同一套 Agent 流程给所有模型复跑 Terminal-Bench 4.0。" }, { "key": "livebench-coding", "name": "LiveBench · 编程综合", "operator": "LiveBench", "url": "https://livebench.ai/#/?cats=Coding%2CAgentic+Coding&ft=1", "license": "Apache 2.0", "status": "ranked", "description": "Coding + Agentic Coding 两项均分，各占 50%。" }, { "key": "artificial-analysis-scicode", "name": "SciCode · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/scicode", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "把物理、化学、生物等科研问题写成能运行的计算代码。" }, { "key": "vals-migration", "name": "Code Migration", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/code-migration", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "把整套程序从一种语言或框架迁移到另一种。" }, { "key": "vals-program", "name": "ProgramBench", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/programbench", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "只给可执行文件，从零重写出行为相同的程序。" }, { "key": "vals-vibe-code", "name": "Vibe Code Bench v1.1", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/vibe-code", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "从需求搭建网页应用，检查实际功能。" }, { "key": "vals-vcb-1-100", "name": "Vibe Code Bench 1–100", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/vcb-1-100", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "在已有应用上连续开发，同时保持旧功能。" }, { "key": "vals-cyber-patch", "name": "CyberBench Patch v1.1", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/cyber", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "修补真实开源项目的安全漏洞。" }, { "key": "deepswe-v1-1", "name": "DeepSWE v1.1", "operator": "DataCurve", "url": "https://deepswe.datacurve.ai/", "license": "官方公开结构化结果；仅管理员私有观察，公开前复核许可", "status": "observing", "description": "在统一的编程助手环境里改仓库、修缺陷，比较的是写代码的模型本身。" }, { "key": "epoch-mirrorcode", "name": "MirrorCode", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/mirrorcode", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "observing", "description": "长时间的软件复现" }, { "key": "taptap-maker", "name": "TapTap Maker", "operator": "TapTap Maker", "url": "https://maker.taptap.cn/leaderboard/", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以易玩／TapTap 条款为准", "status": "reference_only", "description": "在真实游戏引擎里用 Lua 写出能跑起来的功能。分数由引擎执行判定，不用另一个模型打分。" }, { "key": "terminal-bench-4", "name": "Terminal-Bench 4", "operator": "Harbor / Terminal-Bench", "url": "https://www.tbench.ai/", "license": "Apache 2.0 · 官方 harbor-framework/terminal-bench 仓库内的成绩提交；保留署名、协议与修改说明。", "status": "reference_only", "description": "保留模型搭配不同 Agent 在终端任务中的原始成绩，供交叉核对。" }, { "key": "hyper-tau", "name": "Hyper-τ-bench", "operator": "Sierra", "url": "https://sierra.ai/blog/hyper-t-bench-evaluating-agents-that-build-agents", "license": "待确认榜单数据使用边界", "status": "observing", "description": "根据业务资料构建并测试能工作的客户服务代理。" }, { "key": "swe-rebench", "name": "SWE-rebench", "operator": "Nebius", "url": "https://swe-rebench.com/", "license": "官方 Hugging Face CC BY 4.0 数据集是任务题库；尚未确认网页最终成绩的稳定导出与再展示边界。", "status": "observing", "description": "持续收集真实仓库问题，比较模型在多种编程语言中的软件修复能力。" }, { "key": "lhtb", "name": "LHTB", "operator": "Tencent HY / Lehigh 等", "url": "https://zli12321.github.io/LHTB/leaderboard.html", "license": "Apache 2.0 · 官方 IntelligenceLab/LHTB-leaderboard 成绩数据集；保留署名和变更说明。", "status": "observing", "description": "在较长时间内持续使用终端，完成复杂工程与工具任务。" }, { "key": "frontiercode", "name": "FrontierCode", "operator": "Cognition", "url": "https://cognition.com/frontiercode", "license": "待确认榜单数据使用边界", "status": "observing", "description": "真实仓库任务的质量、测试与修改范围" }, { "key": "frontierswe-v2", "name": "FrontierSWE v2", "operator": "Proximal Labs", "url": "https://www.frontierswe.com/", "license": "待确认榜单数据使用边界", "status": "observing", "description": "长时间运行的工程实现与研究任务" }, { "key": "weirdml", "name": "WeirdML", "operator": "WeirdML", "url": "https://htihle.github.io/weirdml.html", "license": "待确认榜单数据使用边界", "status": "observing", "description": "在陌生数据上实现机器学习方案" }, { "key": "artificial-analysis-hle", "name": "HLE · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/humanitys-last-exam", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "横跨学科的专家级难题，不用工具作答。" }, { "key": "epoch-frontiermath", "name": "FrontierMath v2 · Tiers 1–3", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "ranked", "description": "研究级数学推理" }, { "key": "livebench-reasoning", "name": "LiveBench · 推理与数学", "operator": "LiveBench", "url": "https://livebench.ai/#/?cats=Reasoning%2CMathematics&ft=1", "license": "Apache 2.0", "status": "ranked", "description": "Reasoning + Mathematics 两项均分，各占 50%。" }, { "key": "artificial-analysis-critpt", "name": "CritPt · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/critpt", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "研究级物理问题，考验专业推导。" }, { "key": "epoch-chess", "name": "Chess Puzzles", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/chess-puzzles", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "ranked", "description": "根据文字棋局寻找正确走法" }, { "key": "epoch-mystery", "name": "Mystery Game Puzzles", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/mystery-game-puzzles", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "ranked", "description": "从陌生规则推导正确行动" }, { "key": "vals-mysterymechanism", "name": "MysteryMechanism", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/mysterymechanism", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "自己设计实验，去推断一套隐藏的规律。" }, { "key": "vals-terminal-bench-science", "name": "Terminal-Bench Science · Vals", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/terminal-bench-science", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "在终端完成科学工作流与分析。" }, { "key": "artificial-analysis-terminal-bench-science", "name": "Terminal-Bench Science · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/terminal-bench-science", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "在终端完成科学工作流与分析。" }, { "key": "epoch-frontiermath-tier4", "name": "FrontierMath v2 · Tier 4", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/frontiermath-tier-4-v2", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "reference_only", "description": "更高难度的数学研究题" }, { "key": "hle", "name": "Humanity’s Last Exam", "operator": "CAIS / Scale AI", "url": "https://labs.scale.com/leaderboard/humanitys_last_exam", "license": "评测代码 MIT；官方榜单成绩再展示边界尚待确认。", "status": "observing", "description": "横跨学科的高难度学术知识与推理。" }, { "key": "simplebench", "name": "SimpleBench", "operator": "SimpleBench", "url": "https://simple-bench.com/", "license": "样题与评测代码仓库 MIT；官网独立发布的最新成绩脚本再展示边界尚未确认。", "status": "observing", "description": "用日常常识和逻辑问题，检验模型能否识别问题中的实际约束。" }, { "key": "matharena-lean", "name": "MathArena · ArXivLean", "operator": "MathArena / ETH Zurich", "url": "https://matharena.ai/", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "status": "observing", "description": "用 Lean 验证数学证明" }, { "key": "matharena-broken", "name": "MathArena · BrokenArXiv", "operator": "MathArena / ETH Zurich", "url": "https://matharena.ai/", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "status": "observing", "description": "发现和修复数学证明中的错误" }, { "key": "matharena-math", "name": "MathArena · ArXivMath", "operator": "MathArena / ETH Zurich", "url": "https://matharena.ai/", "license": "代码 MIT；数据含 CC BY-NC-SA 4.0 边界，需区分题库与榜单成绩。", "status": "observing", "description": "从近期数学论文提炼的新题" }, { "key": "arc-agi-2", "name": "ARC-AGI-2", "operator": "ARC Prize", "url": "https://arcprize.org/leaderboard", "license": "待确认榜单数据使用边界", "status": "observing", "description": "从陌生规则中学习和泛化" }, { "key": "arc-agi-3", "name": "ARC-AGI-3", "operator": "ARC Prize", "url": "https://arcprize.org/leaderboard", "license": "待确认榜单数据使用边界", "status": "observing", "description": "从陌生规则中学习和泛化" }, { "key": "artificial-analysis-gdpval", "name": "GDPval-AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/gdpval-aa", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "按真实职业交付件出题，两两比较产出质量。" }, { "key": "artificial-analysis-briefcase", "name": "AA-Briefcase", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/aa-briefcase", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "办公知识工作的长任务，两两比较交付结果。" }, { "key": "artificial-analysis-automationbench", "name": "AutomationBench · AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/automationbench-aa", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "跨办公软件完成一整条业务自动化流程。" }, { "key": "vals-finance-agent", "name": "Vals Finance Agent", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/fabv2", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "金融分析师的日常办公题，看研报、建模这类工作做得怎样。" }, { "key": "mercor-apex-agents", "name": "APEX-Agents 1.1", "operator": "Mercor", "url": "https://www.mercor.com/apex/apex-agents-leaderboard/", "license": "官方公开结果；数据与代码许可不等于榜单再分发授权", "status": "ranked", "description": "投行、咨询和法律里的长任务，看模型当助手能不能把一件完整的工作做完。" }, { "key": "vals-harvey-lab", "name": "Harvey Legal Agent Benchmark", "operator": "Vals AI / Harvey", "url": "https://www.vals.ai/benchmarks/hlab", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "处理法律资料并交付 Word、Excel、幻灯片等可审阅文件。" }, { "key": "mercor-tau3-banking", "name": "τ³-Banking · Mercor", "operator": "Mercor", "url": "https://www.mercor.com/apex/oss-benchmarks/oss-tau-3-banking-leaderboard/", "license": "官方公开结果；题目与评测框架来自 Sierra tau2-bench（MIT）", "status": "ranked", "description": "在统一工具环境里办理银行客服业务：查政策、找产品、改账户。" }, { "key": "livebench-data-analysis", "name": "LiveBench · 数据分析", "operator": "LiveBench", "url": "https://livebench.ai/#/?cats=Data+Analysis&ft=1", "license": "Apache 2.0", "status": "ranked", "description": "表格重排、列类型标注、表连接预测等数据处理任务。" }, { "key": "vals-medcode", "name": "MedCode", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/medcode", "license": "官方公开结果；公开再展示保留官方署名，完整再分发授权仍以 Vals 条款为准", "status": "ranked", "description": "根据住院记录给出医疗诊断编码。" }, { "key": "artificial-analysis-analyst-agent", "name": "AA-Analyst Agent", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/aa-analyst-agent", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "可靠地分析数据并回答专业问题。" }, { "key": "tau-banking", "name": "τ³-Banking", "operator": "Sierra", "url": "https://taubench.com/leaderboard/?benchmark=knowledge", "license": "MIT · Sierra Research tau2-bench 官方仓库；保留来源署名。", "status": "observing", "description": "在统一工具环境中办理银行业务，检验规则理解、客户沟通与任务完成。" }, { "key": "zapier-automation", "name": "AutomationBench", "operator": "Zapier", "url": "https://zapier.com/benchmarks", "license": "待确认榜单数据使用边界", "status": "observing", "description": "跨办公软件完成自动化工作" }, { "key": "officeqa-pro", "name": "OfficeQA Pro", "operator": "Databricks", "url": "https://github.com/databricks/officeqa", "license": "代码 Apache 2.0；题目数据 CC BY-SA 4.0，现迁至需要申请访问的 Hugging Face 数据集。", "status": "observing", "description": "从大量真实财务文件中检索、计算并回答问题。" }, { "key": "mcp-atlas", "name": "MCP Atlas", "operator": "Scale AI", "url": "https://labs.scale.com/leaderboard/mcp_atlas", "license": "官方环境代码采用 MIT；该许可不自动覆盖实时网页成绩，榜单数据再展示边界待确认。", "status": "observing", "description": "通过真实 MCP 服务检索资料、操作文档与数据库，完成跨软件任务。" }, { "key": "financebenchmark", "name": "FinanceBenchmark", "operator": "FinanceBenchmark", "url": "https://finbenchmark.ai/", "license": "官方仓库代码为 MIT，网站要求保留 FinanceBenchmark 引用；未确认实时成绩完整导出的使用边界，不将代码许可替代榜单数据许可。", "status": "observing", "description": "完成金融定价、资本计算和风控任务，用数值与代码执行结果核验。" }, { "key": "prbench-legal", "name": "PRBench · Legal", "operator": "Scale AI", "url": "https://labs.scale.com/leaderboard/prbench-legal", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "status": "observing", "description": "回答真实法律工作中的复杂问题，检验专业判断和论证质量。" }, { "key": "prbench-finance", "name": "PRBench · Finance", "operator": "Scale AI", "url": "https://labs.scale.com/leaderboard/prbench-finance", "license": "官方题库和评测代码公开；其许可不自动覆盖实时排行榜成绩，最新成绩的稳定导出与再展示边界待确认。", "status": "observing", "description": "回答真实金融决策问题，检验专业判断、推导和风险说明。" }, { "key": "spreadsheetbench-2", "name": "SpreadsheetBench 2", "operator": "SpreadsheetBench / Renmin University / AfterQuery", "url": "https://spreadsheetbench.github.io/", "license": "官方网站页尾声明 MIT；数据集与代码入口公开。", "status": "observing", "description": "完成财务建模、修复工作簿和数据可视化的整套表格工作。" }, { "key": "vending-bench-2", "name": "Vending-Bench 2", "operator": "Andon Labs", "url": "https://andonlabs.com/evals/vending-bench-2", "license": "待确认榜单数据使用边界", "status": "observing", "description": "经营模拟商店一年，处理采购、谈判、库存与定价。" }, { "key": "epoch-ebr", "name": "EBR-bench", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/ebr-bench", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "reference_only", "description": "在长任务中学习新规则并使用记忆" }, { "key": "vals-excel", "name": "Excel · EMB", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/emb", "license": "待确认榜单数据使用边界", "status": "observing", "description": "表格编辑与 Excel 办公" }, { "key": "vals-legal", "name": "Legal Research", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/legal_research", "license": "待确认榜单数据使用边界", "status": "observing", "description": "法律检索和论证" }, { "key": "vals-tax", "name": "TaxAgentBench", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/tax_agent_bench", "license": "待确认榜单数据使用边界", "status": "observing", "description": "税务专业问题" }, { "key": "vals-medscribe", "name": "MedScribe", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/medscribe", "license": "待确认榜单数据使用边界", "status": "observing", "description": "临床记录整理与撰写" }, { "key": "vals-biomystery", "name": "BioMysteryBench", "operator": "Vals AI", "url": "https://www.vals.ai/benchmarks/biomysterybench", "license": "待确认榜单数据使用边界", "status": "observing", "description": "生物学研究推理" }, { "key": "terminal-science", "name": "Terminal-Bench Science", "operator": "Harbor", "url": "https://www.terminal-bench-science.ai/", "license": "待确认榜单数据使用边界", "status": "observing", "description": "在终端里完成科研任务" }, { "key": "artificial-analysis-omniscience", "name": "AA-Omniscience", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/omniscience", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "事实问答，答错倒扣、拒答不扣，看模型知道什么、会不会乱编。" }, { "key": "epoch-simpleqa", "name": "SimpleQA Verified", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/simple-qa-verified", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "ranked", "description": "短问题的事实准确性" }, { "key": "artificial-analysis-lcr", "name": "AA-LCR", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "读很长的文档后，回答需要综合推理的问题。" }, { "key": "livebench-writing", "name": "LiveBench · 语言与指令", "operator": "LiveBench", "url": "https://livebench.ai/#/?cats=Language%2CIF&ft=1", "license": "Apache 2.0", "status": "ranked", "description": "Language + IF 两项均分，各占 50%。" }, { "key": "bullshitbench-v2", "name": "BullshitBench · 错误前提", "operator": "Peter Gostev / BullshitBench", "url": "https://github.com/petergpt/bullshit-benchmark", "license": "MIT · Copyright (c) 2026 Peter Gostev；官方成绩随同仓库发布，无单独数据许可例外。", "status": "ranked", "description": "识别问题里的错误前提，观察模型会不会顺着错误继续编造。" }, { "key": "artificial-analysis-mlcr", "name": "MLCR-AA", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/evaluations/mlcr-aa", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "ranked", "description": "准确、完整且简洁地理解医疗长记录。" }, { "key": "epoch-gpqa", "name": "GPQA Diamond", "operator": "Epoch AI", "url": "https://epoch.ai/benchmarks/gpqa-diamond", "license": "CC BY 4.0 · Epoch AI 自行评测数据", "status": "reference_only", "description": "研究生水平的物理、化学与生物知识" }, { "key": "facts-parametric", "name": "FACTS Parametric", "operator": "Google DeepMind / Kaggle", "url": "https://www.kaggle.com/benchmarks/google/facts-parametric/leaderboard", "license": "Apache 2.0 · 官方 benchmark 与 Score task 页面明确声明；Kaggle 文档提供公共榜单的免登录 JSON 下载。", "status": "observing", "description": "不借助搜索工具回答世界事实问题，检验模型自身知识的准确性。" }, { "key": "arena-text", "name": "Arena Text", "operator": "LMArena", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "license": "CC BY 4.0", "status": "ranked", "description": "真人匿名两两比较回答，看大家更愿意读哪一个。补上选择题看不出的整体体验。" }, { "key": "artificial-analysis", "name": "AA Index", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/", "license": "Artificial Analysis 公开页面数据；展示须显著署名并链接来源，遵守其 Terms of Use", "status": "cross_reference", "description": "把多项当前评测放到同一套环境里复跑，得到一个综合能力分。价格只作参考，不进总分。" }, { "key": "livebench-general", "name": "LiveBench", "operator": "LiveBench", "url": "https://livebench.ai/", "license": "Apache 2.0", "status": "cross_reference", "description": "题库持续更换，答案可以自动核对。用来看模型换题之后还能不能站住。" }, { "key": "arena-webdev", "name": "Arena WebDev", "operator": "LMArena", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "license": "CC BY 4.0", "status": "reference_only", "description": "同样是真人盲选，比的是网页好不好用、能不能交付。" }, { "key": "arena-creative-writing", "name": "Arena 创作盲选", "operator": "LMArena", "url": "https://arena.ai/leaderboard/text/creative-writing", "license": "CC BY 4.0 · Arena 官方 Hugging Face leaderboard-dataset；保留署名、来源链接并说明聚合改动。", "status": "reference_only", "description": "真人盲选故事、诗歌和其他创意表达，观察作品是否吸引读者。" }, { "key": "eq-creative", "name": "Creative Writing v3", "operator": "EQ-Bench", "url": "https://eqbench.com/creative_writing.html", "license": "MIT · EQ-bench-site README 元数据声明", "status": "reference_only", "description": "短篇创作与表达" }, { "key": "eq-longform", "name": "Longform Writing", "operator": "EQ-Bench", "url": "https://eqbench.com/creative_writing_longform.html", "license": "MIT · EQ-bench-site README 元数据声明", "status": "reference_only", "description": "长篇故事的连贯性与完整性" }, { "key": "eq-emotional-v4", "name": "EQ-Bench 4 · 情绪与人际理解", "operator": "EQ-Bench", "url": "https://eqbench.com/", "license": "MIT · EQ-bench-site README 元数据声明", "status": "observing", "description": "在多轮交流中理解人物情绪、隐含需求与不同的沟通偏好。" }, { "key": "lechmazur-short-story", "name": "Short-Story · 短篇小说", "operator": "Lech Mazur", "url": "https://github.com/lechmazur/writing", "license": "未找到明确的榜单数据复用许可；GitHub 公开不等于开放授权。", "status": "observing", "description": "把固定人物、情节与风格要求自然写进一篇完整短篇小说。" }, { "key": "tonebench-scripts", "name": "ToneBench · 文风与脚本", "operator": "Towards AI", "url": "https://benchmark.towardsai.com/", "license": "公开 JSON 可读取，尚未找到明确的数据聚合与再展示条款。", "status": "observing", "description": "按指定作者文风写视频脚本，考查开场、结构、表达与口播节奏。" }, { "key": "tubelab-scriptwriting", "name": "TubeLab · 视频脚本", "operator": "TubeLab", "url": "https://tubelab.net/benchmark/scriptwriting", "license": "未确认稳定成绩导出和公开聚合许可。", "status": "observing", "description": "面向不同视频频道写脚本，比较故事组织、文风、节奏与清晰度。" }, { "key": "ncbench-author-workflow", "name": "NC Bench · 作家工作流", "operator": "Novelcrafter", "url": "https://www.nc-bench.com/", "license": "尚未确认机器成绩与数据复用条款。", "status": "observing", "description": "帮助作家改写、续写、总结、翻译和整理创作想法。" }, { "key": "design-arena-design", "name": "Design Arena · 视觉设计", "operator": "Design Arena / Intelligence", "url": "https://www.designarena.ai/leaderboard/website", "license": "正式 API 数据允许个人及商业展示，须署名和可见官网链接；API 需申请获批。", "status": "observing", "description": "真人盲选网页、界面、图表和幻灯片的设计作品。" }, { "key": "openvibeeval-design", "name": "OpenVibeEval · 网页作品", "operator": "OpenVibeEval", "url": "https://openvibeeval.com/", "license": "未确认稳定成绩导出与公开聚合许可。", "status": "observing", "description": "对照网页作品、可访问性和盲选偏好，观察不同工具下的设计表现。" }, { "key": "svgbench-arena", "name": "SVGBench · 矢量图盲选", "operator": "SVGBench", "url": "https://svgbench.ai/leaderboard", "license": "未找到明确的成绩复用许可。", "status": "observing", "description": "按相同提示盲选 SVG，考查视觉清晰度与图形表现。" }, { "key": "rapidata-svg", "name": "Rapidata SVG", "operator": "Rapidata", "url": "https://www.benchmark.ai/", "license": "CC BY 4.0 · Rapidata/svg-benchmark；保留官方署名与数据链接。", "status": "observing", "description": "真人比较模型生成的矢量图，直接评价视觉偏好。" }, { "key": "arena-vision", "name": "Arena Vision", "operator": "LMArena", "url": "https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset", "license": "CC BY 4.0", "status": "reference_only", "description": "真人盲选带图片的回答，观察视觉理解的实际体验。" }, { "key": "roboflow-vision", "name": "Roboflow Vision Evals", "operator": "Roboflow", "url": "https://playground.roboflow.com/evals", "license": "待确认榜单数据使用边界", "status": "observing", "description": "读图、计数、OCR 与图像信息提取" }, { "key": "artificial-analysis-multilingual", "name": "AA 中文／多语言", "operator": "Artificial Analysis", "url": "https://artificialanalysis.ai/models/multilingual", "license": "API 数据展示或分享须显著署名；再分发权与定制条款须另行取得并遵守 Terms of Use", "status": "observing", "description": "单独看中文和多语言表现，不再把英文综合能力重复算一次。" }, { "key": "rws-mgate", "name": "RWS M-GATE", "operator": "RWS", "url": "https://www.rws.com/artificial-intelligence/train-ai-data-services/research/m-gate/leaderboard/", "license": "待确认榜单数据使用边界", "status": "observing", "description": "含中文的 30 种语言，考查语法与翻译" } ], "provenance": { "repository": "https://github.com/KKKKhazix/AIHOT", "commit": "3343fe2b20db4be7269113752d82d3992fc52b6b", "registry_sha256": "6cccc872adce3b50b03c10b416925a09a6dab517b806e41454e684a4f3a2a016", "registry_url": "https://github.com/KKKKhazix/AIHOT/blob/3343fe2b20db4be7269113752d82d3992fc52b6b/packages/backend/src/leaderboard/source-registry.json", "method_url": null, "method_source_status": "unavailable_at_pinned_commit", "rules_url": "https://aihot.news/leaderboard/rules", "sources_url": "https://aihot.news/leaderboard/sources" } }