# Start the proxy (uses default Chutes backend)
docker-compose up -d
# View logs
docker-compose logs -f
# Stop the proxy
docker-compose downSet environment variables before running docker-compose up:
# Backend URL (default: https://llm.chutes.ai/v1/chat/completions)
export BACKEND_URL=https://your-backend.com/v1/chat/completions
# Log level (default: info)
export RUST_LOG=debug
# Caddy TLS configuration (default: true)
export CADDY_TLS=false # Set to false for plaintext HTTP
export CADDY_DOMAIN= # Your domain (optional)
export CADDY_PORT=8180 # Caddy port (default: 8180)Or create a .env file:
BACKEND_URL=https://llm.chutes.ai/v1/chat/completions
RUST_LOG=info
# Caddy configuration (optional)
CADDY_TLS=true
CADDY_DOMAIN=
CADDY_PORT=8180docker-compose up -dBACKEND_URL=http://localhost:8000/v1/chat/completions docker-compose up -dRUST_LOG=debug docker-compose up -d# Build the image
docker build -t claude-openai-proxy .
# Run the container
docker run -d \
-p 8080:8080 \
-e BACKEND_URL=https://llm.chutes.ai/v1/chat/completions \
-e RUST_LOG=info \
--name claude-proxy \
claude-openai-proxyThe container includes a health check that runs every 30 seconds:
# Check container health
docker-compose ps
# Manual health check
curl -f http://localhost:8080/healthdocker-compose logs -f claude-proxy