Quick Start

Welcome to BitRouter

BitRouter is the agentic proxy for modern agent runtimes. It's a high-performance Rust proxy that lets LLM agents autonomously discover and route to language models, tools, and other agents — with agent-native control, guardrails, and real-time observability through CLI and TUI interfaces.

Multi-provider routing across OpenAI, Anthropic, Google, and custom providers. MCP gateway for tool discovery. ACP gateway for agent-to-agent dispatch. Agent firewall for inspecting, redacting, or blocking risky content. Agentic payments with 402/MPP handling. All in a single binary with zero infrastructure dependencies.

Agent Skills (Recommended)

The fastest way to get started is to install BitRouter Agent Skills into your agent. Once installed, your agent handles everything — installation, key detection, configuration, and server startup — autonomously.

npx skills add bitrouter/agent-skills

Then just paste this prompt into your agent:

Set up BitRouter as my local LLM proxy. Install it, ask me whether I want
to use the default provider (BitRouter Cloud) or bring my own API keys,
then configure accordingly, start the server, and verify it's working.

Quick Start

Install and launch BitRouter:

curl -fsSL https://install.bitrouter.ai | sh

npm install -g bitrouter

brew install bitrouter/tap/bitrouter

Then start the proxy:

bitrouter

On first launch, the setup wizard walks you through provider configuration. Once running, make a request:

curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-4o",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Point any agent runtime at BitRouter — for example, with OpenClaw:

export OPENAI_BASE_URL=http://localhost:8787/v1
openclaw

Every request now flows through BitRouter's TUI — you'll see live provider routing, token usage, and cost tracking for each model call.

Next Steps