GCMP

Overview ​

GCMP is a Visual Studio Code extension that expands the model choices available in GitHub Copilot Chat.
Like Unify Chat Provider for Copilot, it acts as a bridge between Copilot’s chat experience and external LLM providers. The big difference is focus: GCMP is built around strong first-class support for Chinese-native model providers, while still supporting OpenAI/Anthropic-compatible endpoints.

If you want Copilot Chat UX but broader model access, GCMP is one of the most practical options right now.

What GCMP Actually Solves ​

GitHub Copilot is excellent as a product, but model and provider flexibility can be limited depending on your region, budget, or preferred vendors. GCMP addresses that by letting you:

• Use multiple model providers inside the Copilot Chat workflow
• Bring your own API key per provider
• Add OpenAI-compatible and Anthropic-compatible custom endpoints
• Switch models quickly from the model picker without leaving VS Code

In short: same chat workflow, many more model backends.

Built-In Provider Coverage ​

Based on the official repository, GCMP includes broad support for providers such as:

• Zhipu AI (GLM series)
• MiniMax
• Moonshot AI (Kimi)
• DeepSeek
• Alibaba Cloud Bailian (Qwen family)
• Volcengine Ark
• Kuaishou StreamLake
• Mthreads

It also supports compatible provider modes so you can connect third-party gateways that follow OpenAI or Anthropic API conventions.

Installation and First Run ​

1. Open VS Code Extensions and install GCMP (vicanent.gcmp).
2. Open GitHub Copilot Chat.
3. In the model selector, choose Manage Models.
4. Pick a provider and configure its API key.
5. Enable your target model and start chatting.

That flow is straightforward, and it keeps everything inside VS Code’s native chat panel.

Why It Feels Powerful in Practice ​

1. Regional and provider flexibility ​

If your day-to-day stack includes domestic providers, GCMP removes a lot of friction. You can keep one editor workflow while choosing models that match your latency, pricing, and availability constraints.

2. Strong customization path ​

GCMP exposes advanced settings for provider overrides and compatible models. That means you can tune:

• Base URL
• Headers
• Model capabilities
• Extra request body parameters

This is especially useful when you rely on custom gateways or provider-specific options.

3. More than just chat ​

GCMP also includes:

• Inline completion modes (FIM and NES)
• Token usage statistics and status bar insights
• Context window occupancy visibility
• AI-assisted commit message generation

So it is not only "chat model switching"; it is closer to a full Copilot-side enhancement layer.

Example Configuration Direction ​

You can configure model behavior through settings.json, for example:

{
  "gcmp.maxTokens": 16000,
  "gcmp.editToolMode": "gpt-5",
  "gcmp.rememberLastModel": true
}

And for custom compatible models, GCMP supports rich provider/model definitions (including custom headers and extra body fields), which makes integration with OpenAI-compatible services much easier.

Good Fit vs. Not a Good Fit ​

GCMP is a strong fit if you:

• Want to stay in Copilot Chat but use many non-default models
• Need Chinese-native providers as first-class options
• Care about deeper control over model routing and parameters

It may be overkill if you:

• Only use one provider and never switch
• Prefer a minimal setup with almost no configuration

Final Thoughts ​

GCMP and Unify Chat Provider target a similar problem space: expanding Copilot’s model ecosystem.
GCMP stands out by combining broad provider support, practical regional coverage, and advanced configuration depth in a single extension.

If your workflow depends on model optionality instead of vendor lock-in, GCMP is worth trying.

What’s Next ​

There are also some other Vibe Coding extensions in VS Code, such as:

• Cline
• Roo Code
• Kilo Code

Simple Usage:

{
  "baseUrl": "https://example.com/v1",
  "apiKey": "YOUR_API_KEY"
}