Reflection's First Open-Weight Model Beam Debuts: 501B-Total MoE With 23B Active, Built for Coding and Agents
baseten · x · 2026-10-16
Baseten is serving Reflection AI's first open-weight model, Beam, on day 0. Key facts:
- Text-only MoE: 501B total parameters, 23B active per token, Apache 2.0 license
- Pretrained on 23.8 trillion tokens, refined with large-scale RL
- Targeted at enterprise frontier workloads: coding, code maintenance, long-horizon multi-step tool use; adjustable reasoning effort to trade off length, performance, and compute cost
- Baseten Model API pricing: $1.30/1M input tokens, $0.13/1M output tokens ($4.30 cached input)
Beam marks Reflection's push into open-weight models with an efficiency-focused coding/agentic play.
More from Models
- Microsoft's MAI-Transcribe now handles 61% of all transcription on OpenRouter — mustafasuleyman · 2026-10-16
- HotDogBench tests 10 LLMs on 1,000 hot dog photos — all within each other's error bars — josh_wills · 2026-10-16
- TabPFN sweeps Kaggle: now in 17% of public notebooks and 3 of September's top-10 solutions — ph_singer · 2026-10-16
- Gemini 4 Argon lands in Perplexity, claims new SoTA on WANDR benchmark — AravSrinivas · 2026-10-16
- DL Weekly #477: Haiku 5.5, building a coding agent from scratch, self-compacting agents paper — dl_weekly · 2026-10-16
- Dev says today's DLP tools are all regex-based, is rebuilding with LLM NER — HankYeomans · 2026-10-16