Reflection announces Beam, a 501-billion-parameter open-weight model; weights are still pending
Beam targets coding, reasoning and agentic work. Benchmarks and compute-efficiency figures are company-reported; final red-teaming continues, with an early-access waitlist and no public weights yet.
Reflection announced Beam on October 5: a sparse mixture-of-experts model with 501 billion total parameters and 23 billion active per token, aimed at coding, reasoning and agentic work. It reports pretraining on 23.8 trillion tokens and four weeks of reinforcement learning on 10,500 NVIDIA GB300 GPUs, producing over 100 million rollouts.
Its company-produced benchmark table leaves multiple competitor results unreported; this report obtained no independent reproduction evidence. Compute-efficiency estimates use active parameters and output tokens, excluding prefill, context-dependent attention and serving overhead; they are not user prices. Beam remains in final red-teaming, with an early-access waitlist. Reflection plans to release weights under Apache 2.0 later this month, alongside a technical report, model card and developer tools. Those materials are not yet public, and the announcement gives no usage price.
Sources
- Introducing Beam: Reflection’s 501B open-weight model — Reflection AI