Beam

Reflection AI Launches Beam: 501B-Parameter Open Model Takes on DeepSeek and Qwen

Nvidia-backed Reflection AI released its first open-weight model Beam on October 5: 501B total parameters with only 23B active per token, 1M-token context. The company claims reasoning performance close to Zhipu's GLM-5.2 at one-quarter to one-third the inference compute, with weights coming later in October under Apache 2.0.