Anthropic API Proxy for Gemini, OpenAI & OpenRouter Models 🔄

Use Anthropic clients (like Claude Code) with Gemini, OpenAI, or OpenRouter backends. 🤝

A proxy server that lets you use Anthropic clients with Gemini, OpenAI, or OpenRouter models via LiteLLM. 🌉

Quick Start ⚡

Prerequisites

• OpenAI API key 🔑
• Google AI Studio (Gemini) API key (if using Google provider) 🔑
• OpenRouter API key (if using OpenRouter provider) 🔑
• uv installed.

Setup 🛠️

1. Clone this repository:

   git clone https://github.com/1rgs/claude-code-openai.git
   cd claude-code-openai

2. Install uv (if you haven't already):

   curl -LsSf https://astral.sh/uv/install.sh | sh

   (uv will handle dependencies based on pyproject.toml when you run the server)

3. Configure Environment Variables: Copy the example environment file:

   cp .env.example .env

   Edit .env and fill in your API keys and model configurations:

   • ANTHROPIC_API_KEY: (Optional) Needed only if proxying to Anthropic models.
   • OPENAI_API_KEY: Your OpenAI API key (Required if using the default OpenAI preference or as fallback).
   • GEMINI_API_KEY: Your Google AI Studio (Gemini) API key (Required if PREFERRED_PROVIDER=google).
   • OPENROUTER_API_KEY: Your OpenRouter API key (Required if PREFERRED_PROVIDER=openrouter).
   • OPENROUTER_API_BASE: Your OpenRouter endpoint (Optional, defaults to "https://openrouter.ai/api/v1").
   • PREFERRED_PROVIDER (Optional): Set to openai (default), google, or openrouter. This determines the primary backend for mapping haiku/sonnet.
   • BIG_MODEL (Optional): The model to map sonnet requests to.
   • SMALL_MODEL (Optional): The model to map haiku requests to.

   Default Model Mappings:

   • OpenAI: gpt-4.1 (BIG) and gpt-4.1-mini (SMALL)
   • Google: gemini-2.5-pro-preview-03-25 (BIG) and gemini-2.0-flash (SMALL)
   • OpenRouter: anthropic/claude-3-opus (BIG) and anthropic/claude-3-haiku (SMALL)

   Custom API Endpoints:

   • You can also specify custom API endpoints for each provider if needed:
     • OPENAI_API_BASE
     • ANTHROPIC_API_BASE
     • GEMINI_API_BASE

4. Run the server:

   uv run uvicorn server:app --host 0.0.0.0 --port 8082 --reload

   (--reload is optional, for development)

Using with Claude Code 🎮

1. Install Claude Code (if you haven't already):

   npm install -g @anthropic-ai/claude-code

2. Connect to your proxy:

   ANTHROPIC_BASE_URL=http://localhost:8082 claude

3. That's it! Your Claude Code client will now use the configured backend models through the proxy. 🎯

Claude Code Guidelines

This repository includes a CLAUDE.md file that provides guidance for using Claude Code with this project. It covers:

• Build & run commands
• Test commands with options
• Code style guidelines
• Development best practices

Model Mapping 🗺️

The proxy automatically maps Claude models to either OpenAI, Gemini, or OpenRouter models based on your configuration:

Claude Model	OpenAI Default	Gemini Default	OpenRouter Default
haiku	openai/gpt-4.1-mini	gemini/gemini-2.0-flash	openrouter/anthropic/claude-3-haiku
sonnet	openai/gpt-4.1	gemini/gemini-2.5-pro-preview-03-25	openrouter/anthropic/claude-3-opus

Supported Models

OpenAI Models

The following OpenAI models are supported with automatic openai/ prefix handling:

• o3-mini
• o1
• o1-mini
• o1-pro
• gpt-4.5-preview
• gpt-4o
• gpt-4o-audio-preview
• chatgpt-4o-latest
• gpt-4o-mini
• gpt-4o-mini-audio-preview
• gpt-4.1
• gpt-4.1-mini

Gemini Models

The following Gemini models are supported with automatic gemini/ prefix handling:

• gemini-2.5-pro-preview-03-25
• gemini-2.0-flash

OpenRouter Models

OpenRouter models are supported with automatic openrouter/ prefix handling.

Model Prefix Handling

The proxy automatically adds the appropriate prefix to model names:

• OpenAI models get the openai/ prefix
• Gemini models get the gemini/ prefix
• OpenRouter models get the openrouter/ prefix
• The BIG_MODEL and SMALL_MODEL will get the appropriate prefix based on the provider

For example:

• gpt-4o becomes openai/gpt-4o
• gemini-2.5-pro-preview-03-25 becomes gemini/gemini-2.5-pro-preview-03-25
• anthropic/claude-3-opus becomes openrouter/anthropic/claude-3-opus when using OpenRouter

Customizing Model Mapping

Control the mapping using environment variables in your .env file or directly:

Example 1: Default (Use OpenAI) No changes needed in .env beyond API keys, or ensure:

OPENAI_API_KEY="your-openai-key"
GEMINI_API_KEY="your-google-key" # Needed if PREFERRED_PROVIDER=google
# PREFERRED_PROVIDER="openai" # Optional, it's the default
# BIG_MODEL="gpt-4.1" # Optional, it's the default
# SMALL_MODEL="gpt-4.1-mini" # Optional, it's the default

Example 2: Prefer Google

GEMINI_API_KEY="your-google-key"
OPENAI_API_KEY="your-openai-key" # Needed for fallback
PREFERRED_PROVIDER="google"
# BIG_MODEL="gemini-2.5-pro-preview-03-25" # Optional, it's the default for Google pref
# SMALL_MODEL="gemini-2.0-flash" # Optional, it's the default for Google pref

Example 3: Use OpenRouter

OPENROUTER_API_KEY="your-openrouter-key"
OPENAI_API_KEY="your-openai-key" # Needed for fallback
PREFERRED_PROVIDER="openrouter"
# BIG_MODEL="anthropic/claude-3-opus" # Optional, it's the default for OpenRouter pref
# SMALL_MODEL="anthropic/claude-3-haiku" # Optional, it's the default for OpenRouter pref

Recent Updates

OpenRouter Provider Support (fa2a9b7)

• Added OpenRouter as a supported provider with automatic model mapping
• Updated LiteLLM dependency to version 1.67.4
• Enhanced error handling and logging for better troubleshooting
• Added OpenRouter configuration in .env.example and model mapping logic

Model Mapping Logic Improvement (c01ded7)

• Updated PREFERRED_PROVIDER default to empty string for more intuitive behavior
• Adjusted model mapping logic to better handle provider selection
• Improved error messages to include model name for easier debugging

Testing Framework Enhancements (1c77b3b, cb0fc87)

• Removed deprecated streaming tests and adjusted server behavior
• Updated test scenarios to use latest Claude models
• Added test for claude-small model with the calculator tool
• Added filtering option to test specific scenarios
• Better validation of test results with detailed statistics

Claude Code Integration (cb0fc87)

• Added CLAUDE.md with guidance for Claude Code users
• Includes build & run commands, test options, and code style guidelines
• Formatted for easy use with claude.ai/code

Running Tests

The project includes a comprehensive test suite to verify functionality:

# Run all tests
python tests.py

# Run only simple tests (no tools)
python tests.py --simple

# Skip streaming tests
python tests.py --no-streaming

# Only run tests with a specific name
python tests.py --filter "calculator"

Contributing

Contributions are welcome! Please follow the code style guidelines in CLAUDE.md when submitting changes.