November 13, 2026
How to transfer motion from a video to a photo?
Use the video for the movement and the photo for the person. In Genjutsu AI you upload a clip you already have, then add one photo for each person you want to replace. The photo takes that person's place. The motion, the camera, the cuts, and the original audio stay. The download is 720p. You do not write a prompt.
Start with the video
Drop in mp4, mov, or webm, up to 80MB. A short dance, a walk, or any take where the person is in frame can be the reference. Genjutsu AI splits the file into scenes and finds the people. You do not choose the clip from a template list.
Then add the photo
The photo is who appears. It can be you or someone you know. One photo belongs to one person, and that person stays the same in every shot they appear in. Generate when the photos are in. The steps, timing, camera, cuts, and audio stay with the file you uploaded.
What helps the result
- A clear person in the video
- The filmed body is the movement. A person who is easy to see gives the photo a take to follow.
- A clear photo
- Use a photo of the person you want in the frame. A full-body photo gives the pose more to match.
- A scene of 15 seconds or less
- A scene longer than 15 seconds cannot be generated. Ten credits cover every 15 seconds.
- One photo per person
- Two people need two photos. Anyone without a photo stays as filmed.
What it costs
The first order is $6.90 and covers one video up to 15 seconds. After that, the Starter Pack is $9.99. The price shows before you generate. A failed generation returns its credits.
Put the photo in a clip you have
Sign in, drop the video, and add a photo for each person you want to replace. Finished videos stay on the account.