Beam can refer to two things: the open-weight model from Reflection AI, or the Erlang virtual machine, BEAM. Recent coverage centers on Reflection AI releasing its first open model, Beam, on Hugging Face: a 501B-parameter MoE with 23B active per token, a 1M-token context window, and an Apache 2.0 license, aimed at coding and agentic use, with day-0 support in transformers, SGLang, and vLLM. A separate story covers Imp, a port of DSPy to Elixir/BEAM.
Reflection AI CEO says AI risks 'certainly worry me'Reflection AI CEO Misha Laskin discussed the release of the company's open-weight model Beam, its valuation, and reports of Nvidia expanding its stake, while saying the risks of AI 'certainly worry me.'
Reflection releases 501B open model BeamNew US lab Reflection released Beam, its first open model, on Hugging Face. The MoE model has 501B total parameters (23B active per token), a 1M-token context window, and an Apache 2.0 license. Aimed at coding and agentic use, it has day-0 support in transformers, SGLang and vLLM plus a Rust harness.
Imp: A Full Port of DSPy to the BEAMImp is a full port of DSPy to Elixir/BEAM that turns language-model calls into typed functions with signatures, modules, optimizers and agent loops, running agents as supervised OTP processes and auto-optimizing programs against labeled examples.
The A.I. Daily