ChatGLM Router for GitHub Copilot Chat

A VS Code extension forked from Hugging Face's huggingface-vscode-chat project, modified to integrate ChatGLM (with Coding and General endpoints) into GitHub Copilot Chat.

Compatibility

Due to limited testing resources, compatibility testing may not be exhaustive. Please report any issues on GitHub or in the comments section, and I will address them promptly.

AI Assistance Statement

The development of this plugin utilized ChatGLM Coding to complete most of the API adaptation work.

Quick Start

1. Install the ChatGLM Router extension (search for "ChatGLM Router" in VS Code extensions)
2. Open VS Code Copilot Chat interface (Ctrl/Cmd + Shift + A)
3. Click the model picker and click "Manage Models..."
4. Find "ChatGLM Router" and click "Manage ChatGLM Router"
5. Select Provider (GLM Coding) and enter API Key (get one from https://open.bigmodel.cn/)
6. Choose the models you want to add to the model picker

Available Models

ChatGLM Coding (Default)

• Optimized for code generation and programming tasks
• Endpoint: https://open.bigmodel.cn/api/coding/paas/v4

ChatGLM General (Optional)

• For general chat and non-coding tasks
• Endpoint: https://open.bigmodel.cn/api/paas/v4/
• Enable in settings if needed (disabled by default)
• Same models available, optimized for conversational AI

Custom Providers (Coming Soon)

• Custom provider support is under development
• For custom API needs, please refer to OAI Compatible Provider for Copilot

Configuration

API Key

Configure your ChatGLM API key via the command palette:

• Press Ctrl/Cmd + Shift + P
• Run "ChatGLM Router: Manage ChatGLM Router"
• Select "ChatGLM (Coding & General)"
• Enter your API key from https://open.bigmodel.cn/

Clear API Key

Remove your stored ChatGLM API key:

• Run "ChatGLM Router: Clear ChatGLM API Key" to delete the stored API key
• You will need to re-enter your API key to use the extension after clearing it

Model Selection

Models are prefixed with their provider:

• chatglm-coding:glm-4-plus - ChatGLM Coding endpoint (default, recommended for VS Code)
• chatglm-general:glm-4-plus - ChatGLM General endpoint (enable in settings first)

Note: If a ChatGLM provider does not have an API key configured, it will still appear in the model picker. Selecting or using that model will prompt you to enter an API key (when not in silent mode).

Settings

Configure in VS Code Settings under chatglmRouter:

Note: ChatGLM General is disabled by default. Enable it in settings if you need conversational AI capabilities.

Usage Statistics

Track your API usage with built-in statistics:

View Statistics

• Run "ChatGLM Router: Show Usage Statistics" command
• View total requests and tokens per provider
• See detailed per-model usage

Reset Statistics

• Run "ChatGLM Router: Reset Usage Statistics" command
• Clears all stored usage data

Statistics in Output

• Run "ChatGLM Router: Show Statistics in Output" command
• Displays detailed statistics in an output channel

Note: Statistics are stored locally in VS Code's global state and are estimates (4 chars ≈ 1 token).

Development

git clone https://github.com/OrientLuna/ChatGLM-vscode-chat
cd ChatGLM-vscode-chat
npm install
npm run compile

Press F5 to launch an Extension Development Host for testing.

Common Scripts

• Build: npm run compile
• Watch: npm run watch
• Lint: npm run lint
• Format: npm run format
• Test: npm run test
• Package: npm run package (generates .vsix file)

Architecture

• Multi-Provider Design: Supports ChatGLM Coding and ChatGLM General
• Provider Registry: Built-in providers in src/config.ts
• Statistics Tracking: Usage data tracked in src/statistics.ts
• API-First Model List: Fetches latest models from provider APIs
• Streaming Response: SSE-like streaming with tool call support

Troubleshooting

Models not appearing

1. Check that your ChatGLM API key is configured correctly
2. Run "ChatGLM Router: Manage ChatGLM Router" to verify API key
3. Check VS Code developer console for errors (Help → Toggle Developer Tools)

API Errors

1. Verify your API key has the required permissions
2. Check that the selected model is available on your chosen endpoint
3. Ensure you have sufficient API credits/quotas

ChatGLM Coding vs General

• Use ChatGLM Coding for code-related tasks (recommended for VS Code)
• Use ChatGLM General for conversational AI and non-coding tasks
• Enable ChatGLM General in settings: chatglmRouter.enabledProviders → add chatglm-general

Requirements

• VS Code 1.104.0 or higher
• ChatGLM API key from https://open.bigmodel.cn/

License

MIT License © OrientLuna

Support

• Report issues: GitHub Issues
• ChatGLM Documentation: https://open.bigmodel.cn/