All features. All tiers. No credit card. BrainstormRouter is a free research project — everything is available at no cost while we build and test in the open. Planned GA tiers shown below.
Individual developers exploring the platform.
Growing deployments that want every intelligence system.
Compliance-heavy deployments that need dedicated infra.
The things that make BrainstormRouter different don't live behind a tier.
Bring your own provider keys. Pay providers directly at their published rates. BR never marks up model cost.
Works with any OpenAI SDK, LangChain, Vercel AI SDK, CrewAI, LlamaIndex. Drop-in swap.
Adaptive routing on every plan. The router learns which models work best from your first request.
Guardian intelligence headers on every response. You always know cost, model, tier, efficiency.
Automatic provider failover. Down provider → healthy alternative, no code change.
Starter 100 MB · Team 10 GB · Enterprise unlimited. Cache hits return in ~4ms at 100% cost saved.
No. You bring your own API keys and pay providers directly. We charge for the routing and intelligence layer — currently $0 during the research phase.
Yes. BrainstormRouter ships as a Docker image and npm package. All features work self-hosted. Enterprise customers get dedicated infrastructure support.
One API call to /v1/chat/completions or any other completion endpoint. Cache hits count as requests but at a reduced rate. Memory operations are separate.
During research phase, limits are soft. If you hit them, reach out — we'll raise them.
Yes. Cache hits don't call external providers, so you pay zero model cost. Higher cache hit rate directly reduces total spend.
No. Everything is free during the research phase. No credit card. No contracts.
See Thompson sampling learn your workload. Watch cost drop in the Guardian headers.