Reflection publishes Beam benchmark scores ahead of its weight release

Reflection AI's first scorecard puts Beam near or ahead of some open models on selected coding tests and behind others; its compute comparison excludes several costs of serving a model.

By · Published · Updated

Reflection AI releases Beam as a US rival to China's open models — CEO Misha Laskin says Beam can compete on coding and agentic tasks; Reflection plans to publish the model weights and technical report later this month.

Reflection AI's first scorecard puts Beam near or ahead of some open models on selected coding tests and behind others; its compute comparison excludes several costs of serving a model.