Anthropic API Proxy for Gemini & OpenAI Models 🔄

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

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

Quick Start ⚡

Prerequisites

• OpenAI API key 🔑
• Google AI Studio (Gemini) API key (if using Google 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).
   • PREFERRED_PROVIDER (Optional): Set to openai (default), google, or together. This determines the primary backend for mapping haiku/sonnet.
   • BIG_MODEL (Optional): The model to map sonnet requests to. Defaults to gpt-4.1 (if PREFERRED_PROVIDER=openai) or gemini-2.5-pro-preview-03-25.
   • SMALL_MODEL (Optional): The model to map haiku requests to. Defaults to gpt-4.1-mini (if PREFERRED_PROVIDER=openai) or gemini-2.0-flash.

   Mapping Logic:

   • If PREFERRED_PROVIDER=openai (default), haiku/sonnet map to SMALL_MODEL/BIG_MODEL prefixed with openai/.
   • If PREFERRED_PROVIDER=google, haiku/sonnet map to SMALL_MODEL/BIG_MODEL prefixed with gemini/ if those models are in the server's known GEMINI_MODELS list (otherwise falls back to OpenAI mapping).

4. Run the server:

   ./run-server.sh

   (the script sets sensible defaults, including --reload, and can be customised by editing environment variables inside)

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 (defaulting to Gemini) through the proxy. 🎯

Model Mapping 🗺️

The proxy automatically maps Claude models to either OpenAI or Gemini models based on the configured model:

Claude Model	Default Mapping	When BIG_MODEL/SMALL_MODEL is a Gemini model
haiku	openai/gpt-4o-mini	gemini/[model-name]
sonnet	openai/gpt-4o	gemini/[model-name]

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

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
• The BIG_MODEL and SMALL_MODEL will get the appropriate prefix based on whether they're in the OpenAI or Gemini model lists

For example:

• gpt-4o becomes openai/gpt-4o
• gemini-2.5-pro-preview-03-25 becomes gemini/gemini-2.5-pro-preview-03-25
• When BIG_MODEL is set to a Gemini model, Claude Sonnet will map to gemini/[model-name]

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 Specific OpenAI Models

OPENAI_API_KEY="your-openai-key"
GEMINI_API_KEY="your-google-key"
PREFERRED_PROVIDER="openai"
BIG_MODEL="gpt-4o" # Example specific model
SMALL_MODEL="gpt-4o-mini" # Example specific model

How It Works 🧩

This proxy works by:

1. Receiving requests in Anthropic's API format 📥
2. Translating the requests to OpenAI format via LiteLLM 🔄
3. Sending the translated request to OpenAI 📤
4. Converting the response back to Anthropic format 🔄
5. Returning the formatted response to the client ✅

The proxy handles both streaming and non-streaming responses, maintaining compatibility with all Claude clients. 🌊

Contributing 🤝

Contributions are welcome! Please feel free to submit a Pull Request. 🎁

Claude to OpenAI/Gemini/Together.ai Proxy

This proxy service translates Claude API requests to OpenAI, Gemini, or Together.ai API requests, allowing you to use Claude-compatible code with other LLM providers.

Setup

1. Create a .env file with your API keys:

ANTHROPIC_API_KEY=your_anthropic_key
OPENAI_API_KEY=your_openai_key
GEMINI_API_KEY=your_gemini_key
TOGETHER_API_KEY=your_together_key

# Set your preferred provider (openai, google, or together)
PREFERRED_PROVIDER=together

# Choose which models to use
# BIG_MODEL=gpt-4.1
# SMALL_MODEL=gpt-4.1-mini
# TOGETHER_MODEL=meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8

2. Run the server:

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

Using with Together.ai

Setting PREFERRED_PROVIDER=together in your .env file will route Claude requests to Together.ai using the Llama-4-Maverick model by default. You can specify a different Together.ai model by setting the TOGETHER_MODEL environment variable.

Installation with Together.ai Support

Make sure to install the package with Together.ai support:

# If using pip
pip install -e "."

# If using uv
uv pip install -e "."

# Or simply use the run-server.sh script which handles this automatically
./run-server.sh

Configuration

In your .env file:

TOGETHERAI_API_KEY=your_together_api_key
PREFERRED_PROVIDER=together
TOGETHER_MODEL=meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8

Note: Make sure to use TOGETHERAI_API_KEY (not TOGETHER_API_KEY) as this is what LiteLLM expects.

The server will map Claude models to Together.ai as follows:

• Claude Haiku → Together.ai model (from TOGETHER_MODEL)
• Claude Sonnet → Together.ai model (from TOGETHER_MODEL)

Model Format

Together.ai models must be referenced using the format: together_ai/model_name

For example:

together_ai/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8

The together_ai/ prefix is essential - it tells LiteLLM which provider to use.

Available Together.ai Models

Currently, the server is configured to support:

• meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 (default)

Newline Handling

The server includes options to control how newlines are handled in model responses:

Together.ai Newline Handling

By default, the server processes newlines in Together.ai responses for better readability. Literal \n characters in responses are converted to actual newlines.

OpenAI Newline Handling

For OpenAI models, the server converts actual newlines to literal \n characters in the response to ensure proper display in OpenAI clients.

Command Line Options

You can control newline handling behavior using these command-line flags when starting the server:

# Start the server with default newline handling
uvicorn server:app --reload --host 0.0.0.0 --port 8082

# Process newlines in Together.ai responses (enabled by default)
python server.py --process-newlines

# Keep actual newlines in OpenAI responses (don't convert to \n literals)
python server.py --preserve-openai-newlines

# View all available options
python server.py --help

Or use the run-server.sh script with options:

./run-server.sh --process-newlines --debug-rich

Troubleshooting

If you encounter errors like:

• "Unmapped LLM provider for this endpoint"
• Authentication failures
• Rate limit issues

Make sure you:

1. Have a valid Together.ai API key set as TOGETHERAI_API_KEY (not TOGETHER_API_KEY)
2. Installed the package with Together.ai support (pip install -e ".[together]")
3. Are using the correct model format with the together_ai/ prefix
4. Are using a model that's available in your Together.ai plan