Reflection AI's Beam now available on AI Gateway for free
Source
vercel.com
Author
Jerilyn Zheng
Date
Why it matters
A US open weightsA model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.Full definition →reasoning modelA model trained to think — generating extended internal reasoning before answering — trading time and tokens for accuracy on hard problems.Full definition → for coding and agents is callable through AI Gateway at no cost during launch, with a 1M context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition → and adjustable reasoning effort. Engineers can trial it in existing agent tooling immediately.
Key takeaways · AI-distilled
Vercel says the standard reflection/beam slug moves to paid pricing automatically with no downtime when the free period ends, while reflection/beam-free simply stops serving, so the slug you pick decides what happens to your calls.
Beam was previously available in stealth as stealth/cat-jam, and Vercel says it supports structured outputs as well as tool useA model's ability to call external functions — run code, search the web, edit files — instead of only generating text.Full definition →.
To use Beam inside Claude Code, Codex or Cursor, Vercel says to install the latest Vercel CLI, run npx vercel ai-gateway setup, then select either Beam slug in the agent.
Vercel says AI Gateway passes through provider pricing with no markup and charges no platform fee on inference, including Bring Your Own Key requests.
Terms in this piece · Glossary
open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
tool use — A model's ability to call external functions — run code, search the web, edit files — instead of only generating text.
reasoning model — A model trained to think — generating extended internal reasoning before answering — trading time and tokens for accuracy on hard problems.
