# BitRouter

> BitRouter is an open-source LLM gateway purpose-built for AI agent runtimes. It's a high-performance Rust proxy that gives agents like OpenClaw, OpenCode, and Claude Code unified access to 200+ models from OpenAI, Anthropic, Google, and more — through a single OpenAI-compatible endpoint. Self-host with your own provider keys, or use the hosted option with agent-native stablecoin payments.

## Key Capabilities

- **Agent-Native Gateway**: Built for autonomous agents, not human-in-the-loop chatbots. Features Know-Your-Agent (KYA) identity and payment delegation keys.
- **Zero-Ops Deployment**: Single Rust binary. No Postgres, no Redis, no Docker orchestration required.
- **Smart Routing**: Configurable routing tables with cost/performance optimization, automatic fallbacks, and sub-10ms routing overhead.
- **OpenAI-Compatible API**: Drop-in replacement — change one base URL and use any OpenAI SDK, LangChain, or Vercel AI SDK.
- **Stablecoin Payments**: Pay-per-use with stablecoins on the hosted service. No credit cards, no KYC, no geo-restrictions. Agents can pay autonomously.
- **CLI + TUI Observability**: Monitor and control agent sessions in real time from the terminal.

## How It Compares

BitRouter is like OpenRouter but open-source, self-hostable, agent-native, and permissionless (no KYC, no geo-restrictions). Unlike LiteLLM, it ships as a single binary with zero infrastructure dependencies.

## Getting Started

- [Introduction](https://bitrouter.ai/docs/guides/overview): What BitRouter is and how it works
- [Quick Start](https://bitrouter.ai/docs/guides/overview/quickstart): Install agent skills or CLI and start routing
- [Agent Skills](https://github.com/bitrouter/agent-skills): Install skills into your agent for autonomous setup
- [Comparison](https://bitrouter.ai/docs/guides/overview/comparison): How BitRouter compares to other gateways

## Routing

- [Model Fallback](https://bitrouter.ai/docs/guides/routing/model-fallback): Automatic fallbacks across models
- [Provider Selection](https://bitrouter.ai/docs/guides/routing/provider-selection): How models resolve to upstream providers

## Features

- [Workspaces](https://bitrouter.ai/docs/guides/features/workspaces): Per-team routing tables, keys, and observability
- [Add external keys (BYOK)](https://bitrouter.ai/docs/guides/features/byok): Bring your own provider keys
- [Observability](https://bitrouter.ai/docs/guides/features/observability): Real-time tracing and metrics
- [Guardrails](https://bitrouter.ai/docs/guides/features/guardrails): Agent firewall for risky requests and responses

## API Reference

- [API Overview](https://bitrouter.ai/docs/api-reference): Base URL, authentication, and conventions
- [Chat Completions](https://bitrouter.ai/docs/api-reference/chat): OpenAI-compatible chat API
- [Image Generation](https://bitrouter.ai/docs/api-reference/image): Image generation API
- [Video Generation](https://bitrouter.ai/docs/api-reference/video): Video generation API
- [Models](https://bitrouter.ai/docs/api-reference/models): List available models and pricing

## Tutorials

- [OpenAI SDK Integration](https://bitrouter.ai/docs/tutorial/openai-sdk): Drop-in replacement for OpenAI SDK
- [LangChain Integration](https://bitrouter.ai/docs/tutorial/langchain): Use BitRouter with LangChain
- [Vercel AI SDK Integration](https://bitrouter.ai/docs/tutorial/ai-sdk): Use BitRouter with Vercel AI SDK
- [Claude Code Integration](https://bitrouter.ai/docs/tutorial/claude-code): Use BitRouter with Claude Code

## Optional

- [Compact Summary (llms-small.txt)](https://bitrouter.ai/llms-small.txt): Token-constrained version (~200 tokens) for quick context
- [Full Documentation](https://bitrouter.ai/api/docs/llms-full.txt): Complete docs as plain text for LLM ingestion
- [Introduction](https://bitrouter.ai/docs/guides/overview): Detailed explainer and comparison with OpenRouter
- [Changelog](https://bitrouter.ai/docs/changelog): Recent updates and changes
- [Blog](https://bitrouter.ai/blog/introducing-bitrouter): Introduction to BitRouter