The one-click way
- Pick Person or Pet. Choose who dances: a person (you, or a friend who agreed) or your pet.
- Upload one clear photo. One subject, face sharp and well lit — a full-body photo gives the best stage shot.
- Pay once. Pick a pack (from $9.99, no subscription). Person and Pet cost the same.
- Get your video by email. AI puts the subject on the red-lit stage and dances the 14-second routine — usually ready in about 7 minutes. You get a vertical 9:16 MP4.
- Post it with the sound. Upload it to TikTok or Reels and add the sound in the app. In TikTok or Reels tap "Add sound" and pick the sound the Tecktonik dance (Cheba Man) posts use.
The stage, the lights and the crowd are newly generated, so no footage from the original video ends up in your post. Make yours now →
What photo works best
- Face clearly visible, good light
- No group photos — one subject only
- Only yourself, people who agreed, or your own pet — no public figures
- For a person: a full-body photo in your normal clothes — that outfit is what you'll dance in.
- For a pet: the whole animal in frame, face sharp, no other pets in the shot.
The manual way (what actually works)
If you want to build it yourself with AI video tools, the trick is to separate who is dancing from how they move. Every good version we've seen does three things:
- Take the right 14 seconds. The recognisable part runs from the first “point to the sky” to the lean at the end — see the routine, second by second.
- Make a staged still first. An image model puts your subject on the stage — red light, festoon bulbs, crowd with phones — in their own clothes, before any video is made. This fixes the look once.
- Transfer the motion, not the pixels. Feed the still plus the routine to a motion-transfer model, so only the moves and timing come across. Never paste a face onto the original clip.
What we measured
| Route | Subject | Result | Cost · time |
|---|---|---|---|
| Staged still + Kling 3.0 Motion Control (720p) | Person | All 10 checkpoints matched the original's poses; face, glasses and outfit held through the head-wrap | ≈ $1.40 · ~5 min |
| Staged still + depth video + Seedance 2.5 (480p) | Pet | Moves and framing matched; the pet stayed itself (head, ears, fur) but with a human-like body | ≈ $2.40 · ~6 min |
| Kling Motion Control with a pet photo | Pet | Rejected: “No valid characters detected” — it only drives human bodies | $0 |
One render per row on October 10, 2026, AI-generated test subjects, provider list prices. Your results will vary.
The prompts we use
1 · Staged still (image model, 9:16, your photo attached)
A photorealistic vertical film still of the person from the reference photo standing alone on a raised wooden stage, full body visible from head to shoes, centred, feet planted apart, arms relaxed slightly out from their sides, both hands empty (no microphone, nothing held), calm confident face. Keep exactly their face, hair, skin tone, glasses or accessories and their own clothes from the photo — nothing added. Low camera angle: shot like a vertical phone video held up from inside the crowd below the stage edge, looking up at the subject. The whole stage and the subject are bathed in saturated deep red stage light — skin, clothes and fur all read red-tinted, with dark shadows. An outdoor courtyard party at night: a dark navy sky, strings of warm white festoon bulbs running diagonally across the top of the frame, dark brick buildings behind. Deep red stage light washes over the stage and the subject. In the bottom foreground: the edge of a DJ booth with glowing decks and a low hedge of dark leaves and red flowers. A packed crowd fills the lower part of the frame, many holding up glowing phones to film. Photorealistic, natural phone-camera grain. No text, no logos, no watermark.
2 · Motion transfer (still + the 14-second routine)
The person in the image performs the dancer's routine from the video, move for move and on the same timing, alone on the red-lit stage at the night party: arms raised and wrapped around the head, one arm pointing straight up, arms thrown wide with bent wrists, sharp sideways arm sweeps, arms crossed overhead with bent knees, small leans and steps — feet always planted on the stage. Keep their face, hair and their own clothes exactly as in the image in every frame. The crowd below keeps filming with glowing phones and cheering. Handheld vertical phone video, deep red stage light, festoon lights.
For pets, motion-transfer models that only accept human bodies will refuse the photo; you need a depth-guided video model and the line “real paws, never human hands” in the prompt — or switch to Pet on our generator.
Label it, and mind the rights
Label the post as AI-generated. Use photos of yourself, people who agreed, or your own pet — never public figures. The routine is Cheba Man's: credit him when you post. Background: what Tecktonik dance is.