Reflection's Beam activates 23 billion of 501 billion parameters per token
Original titleReflection's Beam becomes the most capable open-weight model built outside China
- Outlet
- The Decoder
- Published
- Importance
- 3/5
Summary
Reflection released Beam, its first open-weight mixture-of-experts model that activates 23 billion of 501 billion parameters per token. It aims to match GLM 5.2 on coding and reasoning while using three to four times less compute.
Original excerpt
Reflection has released Beam, its first open-weight model. The mixture-of-experts system activates just 23 billion of its 501 billion parameters per token and aims to match GLM 5.2 on coding and reasoning while using three to four times less compute. The strategy against Chinese open-weight rivals like Deepseek and Qwen is efficiency over raw performance. The article Reflection's Beam becomes the most capable open-weight model built outside China appeared first on The Decoder.