One API.
80+ models.
Zero framework bloat: one API across OpenAI, Anthropic, Google, xAI and Meta. When a model gets deprecated or repriced, you swap one line and keep shipping.
- Plain HTTP: curl or any HTTP client, no SDK required
- One OpenAPI spec covers every public endpoint
- Source-available, self-hostable, yours to fork
# Create a key under Profile > API Keys, then: export B4M_API_KEY=b4m_live_... # The response is always a server-sent event stream curl -sN -X POST "https://app.bike4mind.com/api/ai/v1/completions" \ -H "Authorization: Bearer $B4M_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "claude-opus-5-5", "messages": [ { "role": "user", "content": "Build me a React component for a todo list" } ] }'
Built on atomics, not stitching.
No LangChain, no N8N, no Haystack. Retrieval, agents, auth and streaming are all built from scratch.
- Upgrades on our timeline, not a framework author's
- Every line auditable, with no framework dependency tree
- Fork it and you get one complete system
Agents that can't run away with your bill
Every agent loops: observe, reason, act, and each run is capped by executions, sub-LLM calls and a USD budget. Long runs become quests you can pause, resume and inspect.
Run it wherever your data has to live.
Each way to run it draws its own boundary, so “where does my data live?” always has a one-line answer. Self-host quickstart →
Hosted
Sign up and call the API with your key. Nothing to install, nothing to run.
Data lives in Bike4Mind's cloud
# Create a key under Profile > API Keys export B4M_API_KEY=b4m_live_...
Local
The whole stack in Docker Compose, from a prebuilt image. No AWS account needed.
Data lives on your machine
docker compose -f compose.selfhost.yaml \ --env-file .env.selfhost up -d
Your AWS
Fork the source and deploy it in your account, run by your team or by ours.
Data lives in your VPC
Ready to build without the glue code?
Up to 10,000 free credits, no card required. Or start from a cookbook recipe →