A proxy service that allows Anthropic API requests to be routed through an OpenAI-compatible URL to access alternative models.

Claude Proxy provides a compatibility layer between Claude Code and alternative models available through OpenRouter or your chosen base URL. It dynamically selects models based on the requested Claude model name, mapping Opus/Sonnet to a configured "big model" and Haiku to a "small model".

Key features:

• FastAPI web server exposing Anthropic-compatible endpoints
• Format conversion between Anthropic and OpenAI requests/responses (see mapping for translation details)
• Support for both streaming and non-streaming responses
• Dynamic model selection based on requested Claude model
• Detailed request/response logging
• Token counting

Model: deepseek/deepseek-chat-v3-0324

• Python 3.10+
• OpenRouter API key
• uv

Create a .env file with your configuration:

OPENAI_API_KEY=<key>
BIG_MODEL_NAME=google/gemini-2.5-pro-preview
SMALL_MODEL_NAME=google/gemini-2.0-flash-lite-001
LOG_LEVEL=DEBUG

See config.py for more configuration options.

CLAUDE_CODE_EXTRA_BODY MAX_THINKING_TOKENS API_TIMEOUT_MS ANTHROPIC_BASE_URL DISABLE_TELEMETRY DISABLE_ERROR_REPORTING

uv run src/main.py

ANTHROPIC_BASE_URL=http://localhost:8080 claude

The proxy server exposes the following endpoints:

• POST /v1/messages: Create a message (main endpoint)
• POST /v1/messages/count_tokens: Count tokens for a request
• GET /: Health check endpoint

LICENSE