Face Swap
Replace the face in a video with a face from a reference image.
Overview
The Face Swap node takes a video and a reference face image and replaces the face in the video using the Roop model via Replicate. Useful for character replacement, privacy preservation, or creative remixing.
Configuration
| Field | Type | Default | Description |
|---|---|---|---|
| Provider | Select | roop | AI model for face swapping (currently only Roop) |
Inputs & Outputs
Inputs:
- Face Image (required) — clear photo of the target face to apply
- Video (required) — source video containing the face to replace
Outputs:
- Face-swapped video
Credit Cost
| Provider | Credits |
|---|---|
| roop | 13 credits |
Best Practices
- Use a clear, front-facing face photo for best swap accuracy
- Source video should have a clearly visible face
- Shorter videos produce faster results and lower credit usage
- Well-lit, high-contrast face images yield cleaner swaps
Common Use Cases
- Replace a placeholder actor with a final character in a storyboard video
- Anonymise faces in recorded footage
- Apply a consistent character face across multiple generated clips