Provides the capability to perform face swaps directly on videos sourced from YouTube using the Magic Hour AI platform.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Magic Hour MCP ServerFace swap my photo onto this YouTube video"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Magic Hour MCP Server
MCP server for Magic Hour AI — face swap videos, face swap photos, lip sync, face detection, and file management.
Tools
Tool | Description |
| Swap a face onto a video (file/URL source) |
| Swap a face onto a YouTube video |
| Swap specific faces using face detection IDs |
| Swap a face onto a photo |
| Detect faces in an image or video |
| Get detected face IDs and URLs |
| Sync lip movements to an audio track |
| Upload a file to Magic Hour storage |
| Check video project status and get download URLs |
| Check image project status and get download URLs |
| Delete a video project |
| Delete an image project |
Setup
Get an API key from magichour.ai/settings/developer
Install dependencies:
npm installBuild:
npm run buildClaude Desktop Configuration
Add to ~/.claude/claude_desktop_config.json:
{
"mcpServers": {
"magic-hour": {
"command": "node",
"args": ["/path/to/magic-hour-mcp/dist/index.js"],
"env": {
"MAGIC_HOUR_API_KEY": "mhk_live_your_key_here"
}
}
}
}Claude Code Configuration
Add to ~/.claude/settings.json:
{
"mcpServers": {
"magic-hour": {
"command": "node",
"args": ["/path/to/magic-hour-mcp/dist/index.js"],
"env": {
"MAGIC_HOUR_API_KEY": "mhk_live_your_key_here"
}
}
}
}Usage Example
Face swap a dance video:
1. Upload your face image: upload_file → get file_path
2. Start face swap: face_swap_video with video URL + face file_path
3. Poll status: get_video_project_status until "complete"
4. Download from the returned URL
Individual face swap:
1. Detect faces: detect_faces on the source video
2. Get results: get_face_detection_result to see all faces
3. Swap specific faces: face_swap_video_individual with face mappingsLicense
MIT