Become the artist in a music video
The performance stays. The face changes. The track is untouched.
GenSwap keeps the vocal track and the mouth timing of the original performance while replacing the performer, so the result still lip-syncs to the song you started with.
GenSwap this clipNo prompt to write
Gemini watches the clip with its audio and writes the shot description itself. You upload a photo and press one button.
3 shots, detected automatically
Every cut is found, each shot is re-shot on its own, and the edit is stitched back in the original order.
The original soundtrack stays
Music and spoken lines are lifted off the source and re-attached to the result, so nothing is re-sung or re-scored.
Identity locked across the clip
The same photo anchors every shot, and the last frame of each shot carries into the next so the face does not drift.
How it works
- 1. Drop in the video and a photo of yourself.
- 2. We read the clip and show you a Scene Report — the cuts, the camera, who is on screen, the price.
- 3. We trace the motion into a de-identified pose track. The original video is never sent to the generation model.
- 4. Every shot is re-shot on Seedance 2.5, stitched, and given its original audio back.