# Rayjutsu

> Recast the people in a video from one photo each. The motion, lip sync, camera and sound stay as filmed.

Canonical: https://raywake.si/explore/rayjutsu
Last updated: 2026-11-04

Type: Video · Tasks: Video editing

Model IDs (API):
- `rayjutsu`

Open in the studio: https://raywake.si/models/rayjutsu

## How to use Rayjutsu

Upload the source video and one clear photo per character. Name the photos in the prompt as @Image1, @Image2 and so on, and say who replaces whom. With an empty prompt the characters change and the scene stays as filmed.

### Example prompt

> Replace the dancer with @Image1. Keep the room as filmed.

## Review your result

Watch the whole video. Faces and outfits should stay the same from the first frame to the last, lips should match the sound and the camera should move as it did in the source.

## Inputs and settings

- Upload the source video as MP4 or MOV, 4–20 seconds long and up to 100 MB. The result keeps its length, motion, lip sync, camera and sound.
- Upload one photo per character, up to 8. Describe only what should change; the movement and timing already come from the video.
- Choose 480p for drafts, or 720p (default) or 1080p for the final video. The price follows the video length in whole seconds and the resolution; check it before you generate.
- Use only photos of people who have agreed to appear in the video.

The price is the quote returned for your exact input before generating; get a current one in the studio or with `POST /v1/quotes`.

## Related

- [BiRefNet V2 Video Background Removal](https://raywake.si/explore/birefnet/v2/video)
- [FLUX.3 Video](https://raywake.si/explore/blackforestlabs/flux-3/edit-video)
- [Bria Video Background Removal](https://raywake.si/explore/bria/video/background-removal)
- [All models](https://raywake.si/explore)