When will Beam's weights be released, and under what license?

Reflection says the full weights are planned for October 2026 under Apache 2.0. The model is still in final red-teaming and evaluation and has not been publicly released yet.

How does Beam compare to leading open models?

Reflection says Beam matches GLM 5.2 at one-third to one-quarter the inference compute and approaches Qwen 3.8-Max on coding and agent tasks. It also says Beam does not yet match the pure capability of frontier open models like Kimi K3.

Who has invested in Reflection?

Reflection has raised $4.6 billion from investors that include Nvidia, Sequoia, and Lightspeed.

Large Language ModelsOpen-Source AIGIGAZINE AI55m agoOct 6, 2026, 13:00 JST

Reflection's Beam open model matches GLM 5.2

Send on LINE

3 Key Points

What happened
Reflection announced Beam, a 501B-parameter open-weight model with 23B active parameters, claiming parity with GLM 5.2 and Qwen 3.8-Max on coding and agent benchmarks while training on 23.8T tokens.
Why it matters
Beam appears to offer comparable results at one-third to one-quarter the inference compute of GLM 5.2, which could lower the cost of running capable open models in-house, according to Reflection's benchmarks.
What to watch
The model is still in final red-teaming and evaluation and has not been released; the full weights are slated for October 2026 under Apache 2.0, so the efficiency claim hinges on independent verification.

WHO IT HITSEnterprise teams building coding and agentic AI workflows on open-weight models, and developers who self-host rather than rent models, are the ones positioned to benefit if Reflection's efficiency claims hold — though the weights are not yet public.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

FAQ

When will Beam's weights be released, and under what license?: Reflection says the full weights are planned for October 2026 under Apache 2.0. The model is still in final red-teaming and evaluation and has not been publicly released yet.
How does Beam compare to leading open models?: Reflection says Beam matches GLM 5.2 at one-third to one-quarter the inference compute and approaches Qwen 3.8-Max on coding and agent tasks. It also says Beam does not yet match the pure capability of frontier open models like Kimi K3.
Who has invested in Reflection?: Reflection has raised $4.6 billion from investors that include Nvidia, Sequoia, and Lightspeed.

GIGAZINE AIRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.