Reflection launches Beam, a 501B-parameter open-weight MoE with 23B active
bhutanisanyam1 · x · 2026-10-06
- Reflection.ai introduces Beam, its first open-weight model: a sparse MoE with 501B total parameters and 23B active, built for coding, reasoning, and agentic workloads.
- Pretrained on 23.8T curated tokens; a high-compute RL run generated 100M+ rollouts on 10.5K NVIDIA GB300 GPUs over 4 weeks.
- Positions Beam at the Western open-weight frontier: competitive with GLM 5.2, approaching Qwen 3.8-Max on coding/agentic tasks, trailing Kimi K3 on raw capability but winning on inference efficiency.
- Weights, tech report, and model card ship later this month after final red-teaming; early access signups open.
Related event: Reflection AI Debuts Beam, a 501B-Parameter Open-Weight Model(20 posts)→
More from Models
- Dev argues Persimmon is the most underrated Western lab release: realistic user simulators may be quietly powering RL environments — willcb · 2026-10-06
- Developer hits Claude's 5-hour usage limit twice in one day — kevinkern · 2026-10-06
- beam called second-best Western model slated to go open-source — willcb · 2026-10-06
- Subscription math: Anthropic plans offer over 5X OpenAI's value vs equivalent API pricing — soumitrashukla9 · 2026-10-06
- SemiAnalysis: Opus 5.5 gives 5x the subscription value of GPT-6.1 Sol as OpenAI halves $200 plan limits — kimmonismus · 2026-10-06
- Bittensor guard model gains 8 F1 points in 4 weeks to near-SOTA via miner attacks — bittingthembits · 2026-10-06