Beam announcement, open model with 501B parameters and 23B active, full weights to be released later this month, undergoing final red-teaming and evaluations, focus on coding and agentic performance
Summary
Beam 发布了一款 5010 亿参数、230 亿活跃参数的开放模型，主打编码与 agentic 性能，声称效率比 GLM 5.2 高 3-4 倍、比西方领先开源模型高 4 倍以上，完整权重将于本月晚些时候在最终红队测试与评估后发布。
Similar Articles
Beam: Reflection's 501B open-weight model
Reflection AI introduced Beam, a 501B-parameter sparse MoE open-weight model (23B active) built for coding, reasoning, and agentic workloads, pretrained on 23.8T tokens and scaled with a high-compute RL run on 10.5K GB300 GPUs. Beam is positioned as advancing the Western open-weight frontier, approaching GLM 5.2 and Qwen 3.8-Max on coding/agentic tasks while using 3–4× less inference compute, with weights and a technical report due later this month.
Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute cost
Reflection AI 发布了其首个前沿开放权重模型 Beam——一个 5010 亿参数（230 亿活跃参数）的文本混合专家模型，宣称在先进推理基准上与 Z.ai 的 GLM-5.2 相当，但推理算力成本低 3-4 倍，旨在为西方提供对抗中国开源模型的低成本选择。
Welcome Beam: a 501B open model you can own, customize, and scale
Reflection 开源发布 Beam 1.0，一个 501B 参数（23B 激活）的 MoE 大模型，支持 1M token 上下文、Apache 2.0 许可，并附带 Rust harness、vLLM/SGLang/transformers 第一天支持以及 SFT 和 RL 训练配方。
Beam
Reflection AI 推出 Beam：一款 5010 亿总参数（230 亿激活）的稀疏混合专家开源模型，面向编码、推理与智能体工作负载。
reflection/Beam-1.0-501B-A23B · Hugging Face
Reflection 发布开源模型 Beam-1.0-501B-A23B，采用稀疏 MoE 架构（501B 总参数/23B 激活参数），支持 1M token 上下文，主打 token 高效推理、软件工程与智能体工具调用能力，提供 BF16、FP8、NVFP4 多种数值精度版本，采用 Apache 2.0 许可。