Claude LMStudio Bridge V2

🚀 Claude-LMStudio-Bridge

A simple Model Control Protocol (MCP) server that enables Claude to communicate with locally run Large Language Models (LLMs) via LM Studio.

🚀 Quick Start

This bridging tool allows Claude to send prompts to models running locally in LM Studio and receive their responses. It's highly useful for the following scenarios:

• Comparing Claude's responses with those of other models.
• Accessing local models specialized for specific tasks.
• Executing queries when the Claude API quota is limited.
• Processing sensitive queries entirely locally.

✨ Features

• Model Communication: Facilitates communication between Claude and local LLMs in LM Studio.
• Multiple Use Cases: Supports various use cases such as response comparison, specialized model access, and local query processing.
• MCP Tools: Provides useful MCP tools for interacting with Claude and local models.

📦 Installation

Prerequisites

• Python 3.8+
• Anthropic Claude with MCP capabilities
• LM Studio running locally
• A local LLM model loaded in LM Studio

Steps

1. Clone this repository:

   git clone https://github.com/infinitimeless/Claude-LMStudio-Bridge_V2.git
   cd Claude-LMStudio-Bridge_V2
   

2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # For Windows users: venv\Scripts\activate
   

3. Install the required packages (choose one method):

   Using requirements.txt:

   pip install -r requirements.txt
   

   Or install dependencies directly:

   pip install requests "mcp[cli]" openai anthropic-mcp
   

💻 Usage Examples

Basic Usage

1. Start LM Studio and load your favorite model.
2. Ensure that LM Studio's local server is running (usually on the default port 1234).
3. Run the bridging server:

   python lmstudio_bridge.py
   

4. In the Claude interface, enable the MCP server and point it to your locally running bridging service.
5. Now you can use the following MCP tools to chat with Claude:
   • health_check: Check if the LM Studio API is accessible.
   • list_models: Get a list of available models in LM Studio.
   • get_current_model: View the currently loaded model.
   • chat_completion: Send a prompt to the current model.

Advanced Usage

Once connected, you can instruct Claude to use the local model:

Claude, please use the LM Studio bridging service to ask the local model: "What's your opinion on quantum computing?"

Claude will use the chat_completion tool to send the query to your local model and display the response.

📚 Documentation

Configuration

By default, the bridging service connects to LM Studio at the address http://localhost:1234/v1. If your LM Studio instance is running on a different port, modify the LMSTUDIO_API_BASE variable in lmstudio_bridge.py.

Troubleshooting

If you encounter dependency issues, try installing them directly:

pip install requests "mcp[cli]" openai anthropic-mcp

For detailed installation instructions and troubleshooting, refer to the Installation Guide.

📄 License

This project is licensed under the MIT License.

🤝 Contribution

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