Face Transform MCP

Face analysis and transforms — keypoints, beautify, age/gender, swap, cartoon, liveness. (Alpha)

Bring AceDataCloud's Face Transform APIs into Copilot Chat. Detect 90+ keypoints per face, beautify portraits, age or de-age, swap perceived gender, face-swap between photos, cartoonize, and detect liveness.

This extension registers the face MCP server with VS Code so GitHub Copilot and any other agent that speaks the Model Context Protocol can call it directly from chat.

Quick Start

1. Install this extension. VS Code registers the face MCP server automatically.
2. Get an API token from Ace Data Cloud → API Keys. New accounts include free trial credit.
3. Open Copilot Chat in agent mode and ask for a image task — VS Code will prompt for the token the first time and store it securely.

The default config talks to the hosted streamable-HTTP endpoint at https://face.mcp.acedata.cloud/mcp — no Python, no uvx, no local install needed.

Example prompts

• "Detect every face in https://example.com/group.jpg and return their keypoints."
• "Beautify https://example.com/me.jpg with smoothing 15 and whitening 25."
• "Swap the face from https://example.com/headshot.jpg onto https://example.com/scene.jpg."

Tool Reference

8 tools available via this server.

Pricing

All Face APIs are currently in Alpha. Free trial credit on sign-up. See full pricing at https://docs.acedata.cloud.

Configuration

This extension contributes the following entry to your VS Code MCP config:

{
  "servers": {
    "face": {
      "type": "http",
      "url": "https://face.mcp.acedata.cloud/mcp",
      "headers": { "Authorization": "Bearer ${input:acedatacloud_api_token}" }
    }
  },
  "inputs": [
    {
      "type": "promptString",
      "id": "acedatacloud_api_token",
      "description": "Ace Data Cloud API token",
      "password": true
    }
  ]
}

VS Code will prompt for the token on first use and persist it in the OS secret store (Keychain / Credential Manager / libsecret).

Alternative: local stdio (no network roundtrip)

If you prefer running the server locally — for offline dev, air-gapped environments, or to pin to a specific PyPI version — install uv and replace your mcp.json entry with:

{
  "servers": {
    "face": {
      "type": "stdio",
      "command": "uvx",
      "args": ["mcp-face-transform"],
      "env": { "ACEDATACLOUD_API_TOKEN": "${input:acedatacloud_api_token}" }
    }
  }
}

uvx will download and run the latest mcp-face-transform on demand.

Alternative: OAuth via Dynamic Client Registration

The hosted endpoint also accepts OAuth 2.1 with DCR. Drop the headers and inputs blocks and VS Code will run the auth flow on first use (redirect URL http://127.0.0.1:33418 or https://vscode.dev/redirect).

Links

• Hosted endpoint: https://face.mcp.acedata.cloud/mcp
• PyPI package: mcp-face-transform
• Source repository: https://github.com/AceDataCloud/FaceTransformMCP
• Ace Data Cloud platform: https://platform.acedata.cloud
• MCP documentation: https://docs.acedata.cloud

License

MIT — see LICENSE.