Source: vercel.com/changelog/reflection-ai-beam-now-available-on-ai-gateway-for-free
Vercel says Beam, previously in stealth as stealth/cat-jam, is a low-latency, US-built open-weight reasoning model for coding and agentic tasks, with weights released under Apache 2.0. It supports tool calling and structured outputs, has a 1M-token context window, and lets users adjust reasoning effort. Only 23B of its 501B parameters are active per token, which Vercel says helps it reach answers sooner than larger open models.
The standard slug, reflection/beam, will move to paid pricing without downtime when the free period ends. The reflection/beam-free slug will stop serving at that point. To use Beam in Claude Code, Codex, Cursor and other coding agents, Vercel recommends running npx vercel ai-gateway setup and selecting the model.
AI Gateway offers a unified API for calling models, tracking usage and cost, and setting retries, failover and routing rules. Vercel says it passes provider pricing through with no markup and charges no platform fee on inference, including Bring Your Own Key requests.
Why it matters if you build agents
Agent builders can try Beam through the same AI Gateway endpoint they already use for other models, with no platform fee on inference. Code pinned to reflection/beam-free will stop working when the free offer ends, so teams using that slug should plan a switch. Code using reflection/beam will continue on paid pricing without a config change.
Get stories like this every weekday
One short email with the day's agent tooling news.