Command Code for Copilot Chat

Pick Command Code models from the Copilot Chat model picker — and keep everything else Copilot already gives you.

Love Command Code's price-performance but don't want to give up GitHub Copilot's agent mode, tool calling, and polished UI? This extension drops every Command Code model straight into the Copilot Chat model selector — with vision, thinking mode, and your own API key.

Why this extension?

• Don't replace Copilot — power it up. No new sidebar, no new chat UI to learn. Just a new set of models in the picker you already use.
• Agent mode, tool calling, instructions, MCP, skills — all of it still works. Copilot's entire stack, now running on any Command Code model.
• Native vision. Command Code serves Claude, GPT, Gemini, Kimi, Qwen, and Grok models with native image input — no proxy required.
• BYOK, pay Command Code directly. Your API key, your bill, your rate limits. Stored in the OS keychain, never on disk.

Features

Every Command Code model in the model picker

The full Command Code Provider catalog is exposed alongside GPT-4o, Claude, and friends in Copilot Chat's model selector — Claude Opus 5, Claude Sonnet 5, GPT-5.5/5.6, Gemini 3.7 Flash, DeepSeek V4 Pro/Flash, Kimi K3, Qwen 3.8 Max, GLM-5.3, Grok 4.6, and dozens more. Switch models mid-chat without losing history.

Thinking effort per model

Each reasoning-capable model exposes a Thinking effort dropdown directly in the Copilot picker:

• Off — disable thinking for fastest responses
• Light — light reasoning for quick edits
• Standard — recommended for everyday use
• Deep — deep reasoning for complex tasks

The setting is sent to the upstream as reasoning_effort and applies only when the model is selected — no global toggle to forget about.

Native vision input

Vision-capable models (Claude Opus/Sonnet 5, GPT-5.4+, Gemini, Kimi K3, Qwen 3.7+, Grok 4.5, and more) receive image attachments directly — no proxy, no description round-trip, no latency tax.

Inherits every Copilot capability

Because this plugs into Copilot's native provider API, you get the full stack for free:

• Agent mode — autonomous multi-step tasks
• Tool calling — file edits, terminal, workspace search, Git, tests
• Instructions & skills — all your .instructions.md, AGENTS.md, and skills just work
• Zero-data-retention — opt in with one setting; requests route only through ZDR-capable upstreams

Secure by default

API key lives in VS Code's SecretStorage (OS keychain on Windows / macOS / Linux). Never in settings.json, never in your Git history.

Zero runtime dependencies

Pure VS Code API + Node.js built-ins. No Python, no Docker, no local proxy server.

Getting Started

Prerequisites

• VS Code 1.116 or later (this extension relies on the same Copilot Chat provider API surface as upstream)
• GitHub Copilot subscription (Free / Pro / Enterprise — the free tier works)
• Command Code subscription that includes API access (GOAT, Pro, Max, Team, or Provider) — see pricing

Installation

1. VS Code — install from the Visual Studio Marketplace (once published).
2. Editors on Open VSX — install from Open VSX (once published).

Usage

1. Run Command Code: Set API Key from the Command Palette (Ctrl+Shift+P).
2. Paste your Command Code API key (from Studio).
3. Open Copilot Chat, click the model picker, pick any Command Code model.
4. Pick a Thinking effort from the dropdown next to the model name.
5. Chat away.

Models

The provider ships the full Command Code catalog, grouped by company. The most popular picks:

The full list (50+ models across 17 providers) is in src/models.ts.

Settings

Thinking effort is configured from Copilot Chat's model picker for each Command Code model.

Example settings.json for a mirror that renames models:

{
  "commandcode-copilot.baseUrl": "https://my-mirror.example.com/v1",
  "commandcode-copilot.modelIdOverrides": {
    "claude-sonnet-5": "claude-sonnet-5-mirror",
    "gpt-5.5": "gpt-5-5-mirror"
  }
}

Example settings.json for zero-data-retention:

{
  "commandcode-copilot.zdr": true
}

License

MIT