Rumpelstiltskin AI puts one adult photo into the built-in reference dance for 220 credits at 480p. Add a second photo to replace the man watching too. The new MP4 does not include the reference audio.
The dance comes from the built-in reference video. This page has no video upload.
My creations
A blank image-to-video box invents new motion from a still. Rumpelstiltskin AI copies the built-in reference dance onto the people in your photos.
Image to Video is the right next stop when you want a different action. Photo to Animation is the right stop when one still should move and you do not want this dance. Rumpelstiltskin AI keeps the built-in reference video.
Play the new MP4 next to the reference video. Reject it when a face, hands, or clothing no longer match your photos.
Hair and the barn can change even when the dance follows the reference.
Rumpelstiltskin AI copies the built-in reference dance onto the one adult in your photo.
The dance comes from the built-in clip in Examples, not from a written prompt.
Rumpelstiltskin AI sends that video with your photos. Image to video cannot do this job, so it is not offered here.
Result: One new MP4 that follows the reference dance.
The dancer is required. The watcher is optional.
One photo runs Wan 2.2 Animate Replace and keeps the man watching. A second photo switches the run to Seedance 2.5, which recasts both men in one pass.
Result: One or two people placed into the barn.
The dance comes from the built-in reference video. This page has no video upload.
Examples plays that video, and every run on this page receives it with your photos.
Result: The same reference dance on the people in your photos.
On Rumpelstiltskin AI, the first photo is the dancer in the red shirt. The second photo is optional and replaces the man watching in the visor.
The barn and the dance stay with the reference video. One photo runs Wan 2.2 Animate Replace and keeps the watcher. Two photos run Seedance 2.5 and recast both men.
A Rumpelstiltskin AI photo is a JPG, PNG, or WebP up to 10MB. Show the head, shoulders, and upper body clearly, one person per photo.
The dancer photo is required. Do not upload a minor, a celebrity, or anyone who has not agreed.
Rumpelstiltskin AI uses three steps.
Upload the dancer → Add the watcher if you want → Generate the clip
Upload one adult photo you can represent in the first slot.
The second slot is optional. Leave it empty for 220 credits at 480p, or add a second adult to recast both men for 770 credits.
Choose 480p or 720p, generate the MP4, and reject it if a face, hands, or clothing drift. The reference video is not the result.
Use the MP4 when you want that person performing this reference dance. Each job still starts from a photo you can represent.
Upload a photo you can represent. Rumpelstiltskin AI uses it as the dancer who follows the reference video.
Add a second adult who agreed. Seedance 2.5 puts them in the visor seat while photo 1 dances.
Play the new file against the reference. Keep it only when the people, hands, and clothing are still recognizable.
Opening the page costs $0. One photo costs 220 credits at 480p. Two photos cost 770 credits at 480p. 720p costs more.
| Choice | Credits | Use it when |
|---|---|---|
| 9 seconds · 480p · one photo | 220 credits | The default. Wan 2.2 Animate Replace puts your photo in the dance. |
| 9 seconds · 720p · one photo | 450 credits | The same run in a sharper file. |
| 9 seconds · 480p · two photos | 770 credits | Seedance 2.5 recasts the dancer and the watcher. |
| 9 seconds · 720p · two photos | 1710 credits | Both people recast in a sharper file. |
Stay on this page when you want this dance copied onto one photo. A Rumpelstiltskin AI run does not include the reference audio. Move to Image to Video for a different action, or Photo to Animation when the still itself is the job.
| Check | This page | Image to Video | Photo to Animation |
|---|---|---|---|
| Starting point | One or two adult photos plus the built-in reference video on this page. | One or more stills and a prompt you write yourself. | One still, without this reference video. |
| What the model does | Copies the dance from the reference onto the uploaded people. | Invents motion from the still. It does not receive this video. | Moves the still. There is no reference dance. |
| Who is in the clip | The dancer photo is required. The watcher photo is optional. | Whoever you describe. The form does not lock this dance. | The person in the still. |
| Reference video | The built-in reference video is sent with your photos. You cannot upload a different dance. | No reference video is sent. | There is no reference dance. |
| Sound | The reference video can be unmuted. A new MP4 does not include that audio or a released song. | Depends on the model you pick there. | A moving still, not this reference video. |
Check the price on the button. One photo costs 220 credits at 480p and 450 credits at 720p. Two photos cost 770 credits at 480p and 1710 credits at 720p.
There is no customer-count claim. Check your photos, the reference video, and the MP4 you can reject.
Last updated October 15, 2026.
Upload one adult photo for the dancer, add a watcher photo if you want one, choose 480p or 720p, and generate. One photo costs 220 credits at 480p.
Yes. The second slot replaces the man watching in the visor. Two photos run Seedance 2.5 and cost 770 credits at 480p. Leave it empty to keep the original watcher.
One photo runs Wan 2.2 Animate Replace. Two photos run Seedance 2.5. The number of photos picks the model; there is no model menu.
No. You can unmute the reference video, but the generated MP4 does not include that audio or a released track. Add sound later only where you have the rights.
No. Every run uses the reference video already loaded in Examples. This page has no video upload.
More video tools
Generate a Rumpelstiltskin AI clip when the photos are yours to use. One photo is 220 credits at 480p. Back to the generator.
