Extract Frames

Extracts frames (images) from a video or GIF URL. Returns an array of base64 images that can be analyzed or stored. Useful for visually understanding a video by passing the frames to a vision model.

Input Schema

Parameter	Type	Description
video_url	string	The valid public URL of the video or GIF to extract frames from.
frame_count	number	Number of frames to extract evenly spaced across the video (default 5, max 10).

REST Endpoint

Currently, this tool does not have a dedicated public REST endpoint, but is available via the MCP server and Assistant tools.

Example MCP Call:


{
  "name": "extract_frames",
  "arguments": {
    "video_url": "https://example.com/video.mp4",
    "frame_count": 5
  }
}

Response:


{
  "success": true,
  "count": 5,
  "frames": [
    "data:image/jpeg;base64,...",
    "data:image/jpeg;base64,...",
    "data:image/jpeg;base64,...",
    "data:image/jpeg;base64,...",
    "data:image/jpeg;base64,..."
  ]
}