Reflection AI Launches Beam, an Open-Weight Model
In brief
Brooklyn-based startup Reflection AI on Monday unveiled Beam, its first frontier open-weight AI model, claiming it matches leading Chinese open models on reasoning benchmarks while using far less computing power.
According to TechCrunch, Beam is a text-only “mixture-of-experts” model with 501 billion parameters in total, of which 23 billion are active for each request. It was trained on 23.8 trillion tokens and supports a context window of one million tokens. Reflection says it is designed for reasoning, coding and agent-style tasks.
Reflection claims Beam performs on par with Z.ai’s GLM-5.2 on advanced reasoning tests and beats leading Western open models while using three to four times less compute to run. TechCrunch stressed that these results have not been independently verified.
“Open-weight” means the trained model can be downloaded and run by others, unlike closed models offered only through an API. Reflection was founded in 2024 by two former Google DeepMind researchers and, according to PitchBook data cited by TechCrunch, has raised about $4.7 billion from investors including Nvidia and Sequoia Capital.
Why it matters
Most of the strongest open-weight models have recently come from Chinese labs. A competitive US-built open model would give companies and governments more choice about where and how they run AI.
Sources
Every ViralScopeHub news item is an original summary written in our own words. We link to the original reporting so you can read the full story from the publisher.