Reflection unveils Beam: a 501B-parameter open-weight MoE model benchmarking between Qwen 3.8 and GLM 5.2
gethackteam · x · 2026-10-06
Reflection announced Beam, its first open-weight model: a sparse MoE with 501B total / 23B active parameters, built for coding, reasoning, and agentic workloads. It claims frontier Western open-weight performance, competitive with GLM 5.2 and approaching Qwen 3.8-Max.
Pretraining used 23.8T curated tokens, plus a 4-week high-compute RL run generating 100M+ rollouts on 10.5K NVIDIA GB300 GPUs. Weights, tech report, and model card ship later this month after final red-teaming. This tweet is gethackteam's repost noting Beam sits between Qwen 3.8 and GLM 5.2 on benchmarks.
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