Google launches Gemini Decision 1, scoring up to 1,000 questions in one pass at $0.10 per million input tokens with no output charges
On 11 November (US time) Google launched the Gemini Decisions API, powered by a new model, Gemini Decision 1, in the Gemini API in Google AI Studio. Rather than writing long answers, it makes closed-set decisions: against one text, image, audio, video or PDF input it evaluates up to 1,000 yes/no, multiple-choice or rubric questions at once and attaches a confidence score to each answer. It costs $0.10 per million input tokens, with output not billed.

Instead of generating text token by token, the model scores the candidate options in a single forward pass, so for contexts up to 16K tokens, end-to-end latency is 150–250 ms and stays nearly flat from 1 to 100 questions per request. The shared input can be up to 256K tokens and is billed once per request, with each question adding only its own instruction and criteria tokens. There are three question types: yes/no, choice among 2 to 255 options, and an ordered rubric score starting at 0. Every answer comes with a 0-to-1 confidence score so code can route ambiguous cases to a fallback or a human reviewer.
Google says Gemini Decision 1 leads its benchmark suite spanning agent routing, reasoning, ranking and multimodal understanding with the lowest calibration error, and that on the shared suite it is nearly 5x cheaper than Gemini 3.5 Flash-Lite through the Interactions API. These results are published by Google. Teams inside Google already use it: Gemini Notebook routes queries to the right model tier and thinking level with up to 40% lower latency for everyday queries, and Stitch’s rapid canvas experiment, CC from Google Labs and an MCP tool safety layer in Search also run on it.
Developers can get an API key in Google AI Studio, with a no-cost tier for prototyping, and follow Python, JavaScript and REST guides. Google says integrations with OpenRouter, LangChain, Vercel, LiteLLM, Pydantic AI, Pipecat, LiveKit and OpenCode are rolling out.