Quick answer
Starrd's Mi Chico template recreates the viral 'Mi Chico' dance to Jason Derulo's official sound from one or two photos. Photo 1 opens alone with a finger point at the camera, photo 2 bursts in and takes the front for an overhead clap, punches at the lens and an arm-roll solo, then photo 1 swings back in front and finishes alone with finger-guns. One photo makes you dance both parts with a copy of yourself. Both dancers look into the camera the whole way, dance in the place from photo 1, and the video runs 14 seconds vertical 9:16 on Seedance 2.5 for 150 credits (about $7.50). People and pets.
The same dance, any subject — tap one to make it with your photo.
One or two photos in, and out comes you and a friend dancing the viral "Mi Chico" routine to Jason Derulo, taking turns in front of the camera, right where your photo was taken. Only have one photo? You dance with yourself. Here it is from one photo of a dog:
Mi Chico
One or two photos. You and a friend (or your pet) trade the spotlight in the viral 'Mi Chico' dance where your photo was taken. 14 seconds, vertical.
What You Get
- The routine, move for move. Copied from the original clip, timing and all.
- Two dancers, one take. The front spot changes hands three times, and each dancer stays themselves every time.
- Your place. You dance in the room or spot from your first photo, not a generic studio.
- Eyes on the camera. Both dancers look straight into the lens the whole way through.
- The Mi Chico audio, so every point, clap and punch lands on the beat.
- 14 seconds, vertical 9:16, ready for TikTok, Reels and Shorts.
What the Trend Actually Is
The Mi Chico dance is a TikTok trend set to Jason Derulo's official "Mi Chico" sound. The clip we copied was posted by @kadirrrx24 on October 9, 2026, and passed 2 million plays in under a week: two friends on a rooftop, one handheld phone, and a routine where they keep stealing the front spot from each other.
Why it suits AI: it needs two people who both know the moves, and a friend willing to film it in one take. Two photos skips all of that. One photo goes further: you get to be both of you.
The Moves
| Time | What happens |
|---|---|
| 0:00 | Dancer 1, close up, points straight at the camera with a grin |
| ~1s | Hands roll over the stomach and chest, shoulders bouncing |
| ~3s | Dancer 2 bursts in from the left and takes the front |
| ~4.5s | Both arms overhead for a clap |
| ~5s | Fists up: punches thrown at the camera |
| ~7s | Dancer 1 steps out; dancer 2's solo, with a hip roll, forearm rolls and arms crossed in an X |
| ~11s | Dancer 1 swings back in front, hands up by the head |
| ~13s | Dancer 2 steps out; dancer 1 finishes alone with finger-guns |
The camera is a handheld phone at chest height, tilted slightly up, in one continuous take with no cuts.
Who goes where: photo 1 opens and closes the video. Photo 2 has the middle: the clap, the punches and the solo. The setting comes from photo 1.
Pets Welcome
- On two legs. Pets dance upright on their hind legs, the whole 14 seconds.
- Real paws. The points, claps and punches are whole paws, never human hands.
- Mix and match. You and your dog, two pets, or one pet dancing with itself, like the demo.
Your Look, Your Place
- Your own clothes. People dance in what they're wearing in their photo. Only lingerie, sleepwear or swimwear gets swapped for a plain outfit.
- Your own hair. Hair length stays as it is in your photo.
- Your setting. The dance happens where photo 1 was taken. No usable background? You get an open rooftop at dusk, like the original.
- Want something else? Director's Notes can change the outfit or the place.
Put the photo with the best background first: that's where you'll dance. A clear, well-lit room or an open outdoor spot makes the best stage, because the dancers move in and out of frame and the background shows a lot.
The Fastest Way
Mi Chico
One or two photos, the full 14-second routine with the Mi Chico audio. People and pets.
One or two photos, 150 credits (about $7.50), no prompt. It renders on Seedance 2.5 at 480p, vertical 9:16.
Build It Yourself
1. Turn the dance into a depth map. Run the 14-second clip through a depth-estimation model (Depth Anything V2 works). The depth video carries the motion and the timing only, so the original dancers' faces and outfits can't leak into your video.
2. Black out the far background. The original rooftop has a distinctive black sawtooth building, and its outline shows up clearly in the depth map. Push everything behind the dancers to black, or the model will try to build that roofline into your place.
3. Make a character sheet for each subject. From each photo, generate a sheet with a front, three-quarter, back and close-up view, plus one big panel of both dancers in photo 1's setting at the clip's camera angle.
A photorealistic vertical still of two dancers standing side by side from the knees up, both looking straight into the camera with big grins, shot on a handheld phone at chest height tilted slightly up. On the left the subject from reference photo 1, pointing at the camera; on the right the subject from reference photo 2, fists up like a boxer. Each wears their own clothes. The setting is the place behind the subject in reference photo 1. No text.
4. Drive it with the depth map. Feed the sheet, the photos and the depth video to a model that copies motion from a reference video (we recommend Seedance 2.5). Say the sheet decides everything you can see and the depth video decides only the choreography, the camera and the timing. Then list the routine beat by beat with the move, the face and who is in front, because a depth map has no faces and the model won't know when to grin. Lay the Mi Chico audio back on at the end.
Common Mistakes
Driving with the raw clip. The original dancers and their rooftop come through. Use the depth map.
Leaving the background in the depth map. The building's roofline is a strong shape, and the model treats it as part of your setting.
No beat list. Without timed expressions, the faces go flat for 14 seconds. Write when each dancer grins, smirks or throws the fierce face.
Losing track of who is who. The dancers trade the front spot three times. Say that each one stays the same person every time they leave and come back, or faces and outfits drift between them.
Human hands on pets. Say "real paws" or the model gives your dog fingers on the punches.
FAQ
How many photos? One or two. One photo means you dance with yourself.
People or pets? Both.
How long is it? 14 seconds, vertical 9:16.
What does it cost? 150 credits, about $7.50.
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 Nicole Kidman Dance AI Video: another viral dance, solo or duo
- How to Make the KATSEYE 'Animal' Dance AI Video: a crew dance for up to three photos
- How to Make the Alors On Danse AI Video: another two-friends video from one viral clip
