VSLLM APIServer

A VSCode extension that converts VSCode Copilot Language Model functionality into a standard OpenAI API interface, supporting third-party application calls.

Features

• 🚀 Converts VSCode's built-in AI functionality into standard OpenAI API interface
• 🔌 Supports third-party applications through standard API calls
• 💡 Compatible with OpenAI API format
• ⚙️ Configurable server port

Limitations

• Cannot process images
• Tools may not work as expected
• System prompt may not work as expected
• API rate limit follows VSCode Github Copilot's restrictions, paid users are recommended for better experience

Installation

1. Search for "VSLLM API Server" in the VSCode Extension Marketplace
2. Click Install
3. Restart VSCode for the extension to take effect

Usage

1. The API server starts automatically after installing the extension
2. Default listening address: http://localhost:10000/v1
3. Available service endpoints:
   • Get model list: GET http://localhost:10000/v1/models
   • Create chat completion: POST http://localhost:10000/v1/chat/completions

Configuration Options

The following options can be configured in VSCode settings:

• vsllm.server.enable: Enable/disable API server (default: true)
• vsllm.server.port: API server listening port (default: 10000)
  • Valid range: 1024-65535
  • Access address: http://localhost:[port]/v1

Development

Requirements

• Node.js
• VSCode

Build Steps

1. Clone the project

git clone [repository-url]
cd vsllm

2. Install dependencies

npm install

3. Compile project

npm run compile

4. Run tests

npm test

Notes

• Ensure the port is not occupied by other applications
• Extension requires VSCode 1.85.0 or higher
• API server starts automatically when VSCode launches

License

LICENSE

Feedback

For issues or suggestions, please submit an issue on the GitHub repository.