Reflection ships Beam, its first open-weight model, after a year of building pretraining from scratch
bhutanisanyam1 · x · 2026-10-06
Reflection AI's alexpolozov announced Beam, the company's first open-weight model, the result of about a year of work. When he joined last November, pretraining was 5-10 people with no crawler, GPU infra, dedup pipeline, in-house evals, or sufficiently large MoEs — all built in a 2026 sprint. He credits a blend of scientific rigor (no YOLO decisions in pretraining) and startup-style speed and ownership.
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