Skip to content

Quick Start

This guide gets Routeplane routing for your agent in under a minute. Routeplane runs as a local proxy — a single binary on your machine that routes LLM requests through your existing subscriptions.

Terminal window
curl -L -O https://routeplaneapp.vercel.app/downloads/routeplane-v0.1.0-source.tar.gz
tar -xzf routeplane-v0.1.0-source.tar.gz
cd johnmwhitman-routeplane-*/
cargo install --path apps/routeplane

(Once Routeplane is published to crates.io, this becomes cargo install routeplane. See Installation for details.)

Login to the providers you subscribe to:

Terminal window
routeplane providers login github-copilot
routeplane providers login xai
routeplane providers login openrouter

Use routeplane providers list to see all available providers and their auth status.

Terminal window
routeplane serve

The daemon starts on http://127.0.0.1:4356. It stays running in the background — open a new terminal for the next step.

Terminal window
curl http://127.0.0.1:4356/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "@fast",
"messages": [{"role": "user", "content": "Hello!"}]
}'

Use model presets (@fast, @standard, @heavy) for automatic best-provider routing, or specify a provider directly like openai/gpt-4o.

Point any OpenAI-compatible runtime at http://127.0.0.1:4356/v1 to route through Routeplane:

Terminal window
# Claude Code
claude --proxy http://127.0.0.1:4356
# OpenCode
opencode --api-base http://127.0.0.1:4356
# Any OpenAI-compatible client
OPENAI_BASE_URL=http://127.0.0.1:4356/v1