Upload a finished clip and a photo of the new hero, and the AI puts them in place of the person on screen. Motion, lighting and background stay from the original video, so one good clip becomes dozens of versions with different faces.
Seedance 2.0 transfers a face cleanly and accepts photos of people, but may reject famous people. Wan 3 accepts any photos of people.
Some models return a silent clip. The “Sound from the clip” button lays the original speech over the result for free, trimmed to the picture or stretched to the sound.
Shoot the clip once and change the hero for each country or audience. No reshoots and no actors needed.
Up to 15 seconds per run. For Seedance the source clip must be at least 4 seconds long.
Some models return the result without sound. The original sound comes back with a separate free step, “Sound from the clip”.
No. You top up a balance and pay only for runs. You can look around without an account; you need to sign up to run something.