What you need before you start
Two photos and enough credits for the resolution you want. The first photo fills the slot closest to the camera, the second fills the slot behind it. Neither photo needs to be posed like the routine, because the pose comes from the template rather than from you.
If you want the pet version, the second slot takes a cat or a dog instead of a person. Everything else in the workflow is identical.
Step 1: pick the front photo
This slot covers the most screen time, so it is worth getting right. Use a recent, front-facing photo with even lighting and one person in frame. The generator takes the face, hair, skin tone, glasses and outfit from this photo, so anything covering the face will carry into the render.
Avoid sunglasses and heavy filters. The template already supplies the expression and the gestures, so a neutral, well-lit portrait gives the most consistent identity across the 22 seconds.
Step 2: pick the second photo
The second slot is the performer standing behind. A clear, front-facing photo of one person works, and so does a clear photo of a cat or a dog.
Keep one subject in the frame and crop to the head and shoulders before uploading. The template carries the pose, so you do not need a full-body shot. If you upload a pet, the generator keeps the animal's own face, fur and body shape rather than turning it into a person.
Step 3: keep the template and set the frame
The scene is fixed. The fence, the ivy, the light, the routine and the camera roll all come from the template, and the ratio controls the output canvas rather than the motion. Leave it on Auto to match the source frame, or pick a fixed ratio when the clip needs to travel to a feed post or an embed.
Resolution is the other decision. 480P is the cheapest way to test an idea, 720P Standard is the practical default for finished social content, and 1080P Pro costs more credits for a higher-detail render.
Step 4: generate and download
Confirm the credit cost, submit the task, and wait for the render. Most tasks finish within a few minutes, and the preview panel shows progress while the provider works.
When the clip comes back, preview it and download the file. The result is a silent 9:16 vertical video with no watermark, which is exactly what you want for TikTok, Reels or Shorts: add the sound inside the app and post.
Common questions
Can I use two photos of the same person? Yes. Both slots are independent, so you can put yourself in front and yourself behind, or two different people, or a person and a pet.
Does it include the song? No. The track is copyrighted, so the render is silent and you add the audio in the app you publish from.
What if the generation fails? The unused credits are refunded automatically, and the failed task stays in your history so you can see what happened.
Try it
Upload two photos, keep the template, and render the routine. If you want to see how the two slots behave with different pairings first, the examples on the landing page cover two human pairs and a fully generated character scene.
Ready to make your own version?
Upload two photos and generate the Smalltown Boy dance clip in a few minutes.