ComfyUITemplates.com

Discover free ready-made ComfyUI templates for AI workflows.

MoCha - Replace Anyone in a Video with ComfyUI

MoCha is a ComfyUI workflow for fast, interactive video character replacement. Upload a video and a reference image, then use green and red canvas markers to select which characters to swap or keep.

MoCha - Replace Anyone in a Video with ComfyUI

ComfyUI Workflow: MoCha – Quick Video Character Replacement

MoCha is a ComfyUI workflow engineered for fast character replacement within video content. It enables users to substitute existing characters in a video with new ones through a simple, canvas-based selection process — no complex masking or manual frame editing required.

How It Works

  • Upload video: Provide the video file where character replacement is needed.
  • New character image: Submit a still image of the replacement character. Using an image with a plain or transparent background improves character recognition accuracy.
  • Interactive canvas selection: Use green circles to mark characters for replacement and a red circle to designate characters to keep.

Quick Start in ComfyUI

  • Load video: Connect your source video to the workflow input node.
  • Input new character: Upload a suitable still image of the character to be inserted.
  • Mark selections: On the canvas, place green circles on characters to replace and a red circle on characters to retain.
  • Process: Execute the workflow to generate the output video with the updated character.

Why Use This Workflow

  • Rapid replacements: Achieve character swaps quickly with minimal setup.
  • User-friendly: Straightforward input and canvas selection stages make it accessible to all skill levels.
  • Visual guidance: Clear colour-coded markers give precise control over which characters are replaced or preserved.

Use Cases

  • Replacing characters or actors in short video clips
  • Creating personalised video content using a reference photo
  • Rapid video prototyping for creative and commercial projects
  • Experimenting with character designs within existing footage