DeepSeek for Copilot

Use DeepSeek V4 models directly in VS Code's Copilot Chat.

But How!?

1. Install this extension
2. Open Command Palette (Ctrl+Shift+P / Cmd+Shift+P)
3. Run DeepSeek: Set API Key
4. Enter your API key from platform.deepseek.com

Usage

Open Copilot Chat and pick a DeepSeek model from the model selector:

• DeepSeek V4 Flash
• DeepSeek V4 Flash Thinking
• DeepSeek V4 Pro
• DeepSeek V4 Pro Thinking

That's it. Chat away.

Requirements

• VS Code 1.108 or later
• GitHub Copilot subscription
• DeepSeek API key

Remote Development

This extension runs on your local machine and works with Remote-SSH, Remote-WSL, and other VS Code remote setups. Network requests to the DeepSeek API are sent from the local client, so the remote host does not need outbound internet access.

DeepSeek API Notes

This extension uses the current DeepSeek V4 API model IDs:

• deepseek-v4-flash
• deepseek-v4-pro

The older deepseek-chat and deepseek-reasoner aliases are deprecated by DeepSeek and scheduled to stop working on 2026-07-24.

Settings

License

MIT