# ProxyLLM Documentation

> ProxyLLM is an OpenAI-compatible gateway. Apps call https://api.proxyllm.ai/v1 like OpenAI; requests route to your Codex subscription, your own Claude Code bridge, or saved provider keys.

Every docs page is also plain Markdown: append `.md` to its URL.

## Guides

- [Get started](https://proxyllm.ai/docs/get-started.md): Point your app at ProxyLLM and make your first request in about five minutes.
- [Connect Codex](https://proxyllm.ai/docs/codex.md): Route requests through the ChatGPT/Codex subscription you already pay for.
- [Claude Code Bridge](https://proxyllm.ai/docs/claude-code.md): Run inference on your Claude subscription through a server you own.
- [Routing keys](https://proxyllm.ai/docs/routing-keys.md): One key that authenticates your apps and carries the routing plan: which providers serve each request, in what order.

## API reference

- [API reference](https://proxyllm.ai/docs/api.md): The base URL, how to authenticate, and every endpoint and error code the gateway serves.
- [Authentication](https://proxyllm.ai/docs/api/authentication.md): One Authorization header. Routing keys authenticate inference; account tokens authenticate management.
- [Agent signup](https://proxyllm.ai/docs/api/agent-signup.md): POST /api/auth?action=agent-signup - autonomous agents create an account with an email OTP and receive an sk_ account token. No captcha, no browser.
- [Chat completions](https://proxyllm.ai/docs/api/chat-completions.md): POST /v1/chat/completions - OpenAI-compatible chat, with streaming, tools, and lane fallback.
- [Responses](https://proxyllm.ai/docs/api/responses.md): POST /v1/responses - the OpenAI Responses API, forwarded natively to Codex and OpenAI lanes.
- [Models](https://proxyllm.ai/docs/api/models.md): GET /v1/models - the gateway's model catalog, including the flagship and mini tiers.
- [Key introspection](https://proxyllm.ai/docs/api/key.md): GET /v1/key - what the routing key in the Authorization header can do, without dashboard access.
- [Blitz](https://proxyllm.ai/docs/api/blitz.md): POST /blitz - run up to 50 prompts through the key's routing in one request.
- [Me](https://proxyllm.ai/docs/api/me.md): GET /v1/organizations/me - who this account token belongs to and what it may do.
- [Usage](https://proxyllm.ai/docs/api/usage.md): GET /v1/organizations/usage - 30-day usage summary: requests, tokens, cost, savings.
- [Routing keys](https://proxyllm.ai/docs/api/routing-keys.md): GET | POST /v1/organizations/routing-keys - list your routing keys or create one.
- [Bridges](https://proxyllm.ai/docs/api/bridges.md): GET | POST /v1/organizations/bridges - list your Claude Code bridges or register one.
- [Codex session](https://proxyllm.ai/docs/api/codex-session.md): GET | POST | DELETE /v1/organizations/codex-session - Codex connection status, connect, disconnect.