Quick answer
Upload one photo to Starrd's Cheba Man template and it puts you, or your pet, alone on a raised stage at a packed outdoor night party under deep red stage light and strings of festoon lights, dancing the viral Cheba Man routine move for move — arms spread wide, both hands thrown up, hand-on-head poses, sharp arm waves, leans and a point to the sky — while the crowd films you on their phones. One continuous 30-second handheld take, vertical 9:16, with the song. You keep your own clothes; pets dance upright with real paws. 200 credits (about $10).
The same dance, any subject — tap one to make it with your photo.
One photo in, and out comes you (or your pet) alone on a stage at a packed night party, red lights on, the crowd filming every move: the Cheba Man routine, 30 seconds, one take. Here's one made from a single photo:
Cheba Man
One photo. You or your pet owning the stage at a night party, the crowd filming. 30 seconds, vertical.
What You Get
- The stage is yours. You stand alone on a raised stage at an outdoor night party, deep red light on you, strings of festoon lights overhead.
- The full routine. Arms spread wide, both hands thrown up overhead, hand-on-head poses, sharp arm waves, leans, steps and a point to the sky, all on the beat.
- The crowd. A packed crowd in front of you, phones up, filming the whole thing.
- The song, so every move lands where it should.
- 30 seconds, vertical 9:16, one continuous handheld take, ready for TikTok, Reels and Shorts.
What the Trend Actually Is
The original is a phone video shot from inside the crowd at an outdoor party at night: one guy on a raised stage, lit red, dancing like nobody else exists while everyone below films him.
Why it works: it's pure confidence. He isn't a trained dancer and doesn't need to be; he owns the stage anyway, and the crowd's phones are the proof.
Why it suits AI: the joke is the attitude, not the dancer. Your dad, your best friend, your dog on its hind legs: the less likely the star, the better it lands.
Pets Welcome
- Up on two legs. Your pet stands upright on the stage like a person.
- Real paws. Its front legs do the arm spreads and waves, with real paws, never human hands.
- Still your pet. Its own face, fur and markings the whole time, nothing it isn't wearing in its photo.
Use one clear, well-lit photo where the face is sharp. The stage, the lights and the crowd come from the template; your photo only decides who's dancing.
The Fastest Way
Cheba Man
One photo, the full 30-second stage routine with the crowd filming. People and pets.
One photo, 200 credits (about $10), no prompt. The template holds the stage, the lights, the crowd, the moves and the song; your photo only decides who's on stage. It renders on Seedance 2.5 at 480p, vertical 9:16.
Build It Yourself
1. Turn the clip into a depth map. Cut the 30 seconds you want and run them through a depth-estimation model (we used Depth Anything V2). You get a greyscale video of shapes, distance and motion only: the routine and the camera survive, but the original dancer's face, outfit and sunglasses don't, so none of them can leak into your video.
2. Make one still of you on the stage. One image of you on the red-lit stage, in your own clothes, with the festoon lights and the crowd.
A photorealistic vertical film still of the subject from the reference photo, standing alone on a raised wooden stage at a packed outdoor night party, mid-dance with both arms spread wide, full body visible. Shot like a phone video from inside the crowd, looking up at the stage. Deep red stage light, strings of warm white festoon lights across the top of the frame, a dark night sky. In the foreground, the crowd holds up glowing phones to film. Keep their exact face and their own clothes. No text.
3. Drive it with the depth map. Feed the still and the depth video to a model that copies motion from a reference video (we recommend Seedance 2.5). Tell it the still decides everything you can see and the depth video decides only the movement, the camera and the timing, and say plainly that the result is in full colour, never grey. Lay the song back on at the end.
Common Mistakes
Using the raw clip. The original dancer's face, striped shirt and sunglasses come through into your video. A depth map carries the moves without any of that.
Forgetting to say "full colour". A model reading a greyscale reference can hand you a grey video. Say it in the prompt.
Letting a pet stand like a person with hands. Say "real paws, never fingers or a human hand", or the model may copy the dancer's hands onto your pet.
FAQ
How many photos? One.
Does it work with my pet? Yes. It stands up on its hind legs and dances with real paws.
How long is it? 30 seconds, vertical 9:16.
What does it cost? 200 credits, about $10.
Related Reading
- How to Make an AI Dance Video: the full guide to dance templates
- How to Make Your Pet Dance: pets on the dance floor
- How to Make the 'Turn The Lights Off' AI Video: another one-take party dance
- How to Make the 'Put A Ring On It' Pet AI Video: the other depth-map dance template
