About the Somebody That I Used to Know AI Video Generator
A two-person painted-wall music-video template inspired by Somebody That I Used to Know. Your two photos supply the identities, hair and clothing; the built-in scene supplies the geometric mural and body paint.
Wiro GPT Image 2.5 first recreates the landscape scene with your two people against a geometric painted wall. The first person faces the camera on the left, and the second looks left in profile on the right. Their faces, hair and clothing follow the uploaded photos, while matte paint is applied only to visible skin. Higgsfield Genjutsu Motion Transfer then animates that completed scene using the source performance, camera motion and original audio. The preview is the source reference, not a generated result. This independent template has no artist affiliation or endorsement.
How to make a Somebody That I Used to Know AI video
- Upload the left person. Upload a clear photo of the person for the left role.
- Upload the right person. Upload a clear photo of the person for the right role.
- Choose 480p or 720p. Choose 480p or 720p and generate the painted-wall performance.
- Preview and download. Preview and download the finished video.
Two photos, two fixed roles
Photo 1 always becomes the person on the left, facing the camera. Photo 2 always becomes the person on the right, in profile looking left. Keep the photos in the intended left/right order: the roles are never swapped.
Your hair and clothes, painted skin
Each person keeps their own face, hair and clothing from their photo. Hair is never painted or recolored, and the clothing stays unpainted. The geometric pink, salmon, burgundy, grey and green paint is applied only to visible skin, with thin soft dividing lines that continue the mural behind them.
How the video is made
First, the painted-wall scene is recreated as a single 16:9 still with both of you in it. Only that finished scene is then animated: the left person follows the frame-left performer and the right person the frame-right performer, frame by frame, with the original audio. Your original photos are not sent to the animation step.
Photo tips for a better result
- Use one person per photo with a visible face and hairstyle.
- Include the clothing you want to keep in each reference.
- Keep the photos in the intended left/right order.
About the preview on this page
The video and poster on this page are the source reference, not a generated result. They show the performance, timing and audio your video will follow.
Independent template
This is an independent AI template. It is not affiliated with, endorsed by or sponsored by Gotye, Kimbra or anyone involved in the original music video. The song title is used only to describe the trend. Only upload photos you have the right to use.
Somebody That I Used to Know video credit cost
| Resolution | Cost |
|---|---|
| 480p | 1,479 credits |
| 720p | 3,022 credits |
The price covers both the scene image and five billed seconds of video at the resolution you choose, and is shown before you generate.
Frequently asked questions
How many photos are required?
Exactly two. Photo 1 becomes the left person and Photo 2 becomes the right person.
Are hair and clothes painted?
The workflow keeps each person's hair and clothing unchanged and applies the geometric paint only to visible skin.
Do I need a reference video?
No. The motion reference is built in, and the output follows its timing and audio.
Is the preview a generated example?
No. It shows the source performance used as the motion and audio reference.
How many credits does it cost?
1,479 credits at 480p or 3,022 credits at 720p, covering the scene image and five seconds of video.
Discover more templates in the trends gallery or explore AI video effects.