← Tuesday, October 6, 2026Models & trainingRelease

Reflection's 501B Beam model used 100 million RL rollouts on GB300s

reflection.ai 17 min read Editor score 7/10

Beam is a 501-billion-parameter sparse model with 23 billion active parameters; its RL run generated over 100 million rollouts on 10,500 GB300 GPUs in four weeks. The scale is useful context for teams evaluating model-training investment, though weights and evaluation details are still forthcoming.

What

Beam is a sparse Mixture-of-Experts model with 501 billion total parameters and 23 billion active, aimed at coding, reasoning, and agentic workloads.

How

It was pretrained on 23.8 trillion tokens; its high-compute RL run generated over 100 million rollouts on 10.5K NVIDIA GB300 GPUs over four weeks.

Why it matters

Beam reports reasoning scores comparable to GLM-5.2 with 3–4× less inference compute, a tradeoff relevant to serving capable models efficiently.

Read on reflection.ai → Hacker News · 131 comments

Headline, summary, brief and diagram are written by a model from the source's own text; the link, byline and numbers come from the source. Read the original before relying on a claim.