Quick answer
Starrd's Closer (Dance) template turns two photos into the viral 'Closer' dance battle from @gnomeboys (12.5 million+ plays). Photo 1 is the one waiting in a multi-storey car park: the film opens behind their head, they take off their sunglasses and stare down the lens. Photo 2 sprints in from the far end, leaps, holds a palm to the lens, shoves photo 1 back by the neck and chest in a staged dance-battle move, and lip-syncs 'I just can't stop' in two close-ups. Photo 1 grins, and the two battle it out on the open rooftop before it loops back to the opening shot. Twenty shots cut where the original cuts, to Ne-Yo's 'Closer' at about 124 BPM. Both keep their own face, hair and outfit. People only. 23 seconds, vertical 9:16, Seedance 2.5, 480p, 250 credits (about $12.50).
The same dance, any subject — tap one to make it with your photo.
Two photos in, and out comes you and a friend in the viral Closer dance battle: one of you waits, the other sprints in, and you battle it out across a sunny concrete car park to Ne-Yo. Here it is:
Closer (Dance)
Two photos. One of you waits, the other sprints in, shoves and sings 'I just can't stop', then you battle it out on the rooftop. 23 seconds.
What You Get
- All twenty shots. The whole 23-second battle, cut where the original cuts, every cut within a couple of frames.
- The story. The wait, the sunglasses coming off, the sprint, the leap, the palm to the lens, the shove, the grin and the rooftop battle.
- The lip sync. Photo 2 sings "I just can't stop" in both close-ups on the original timing.
- On the beat. Ne-Yo's "Closer" at about 124 BPM, every move where the original's lands.
- Your own outfits. Faces, hair and the clothes from your photos, not the original's sweater and track jacket.
- A clean loop. It ends on the same over-the-shoulder shot it opens with.
- 23 seconds, vertical 9:16, with the original sound, ready for TikTok, Reels and Shorts.
What the Video Actually Is
@gnomeboys posted the clip on April 2, 2026, captioned "who's gonna stop us". It passed 12.5 million plays, 1.9 million likes and 177,000 saves. Two friends turn a multi-storey car park into a music video: one waits under the concrete deck, the other sprints at him from the far end, and they square off and dance-battle across twenty quick, handheld shots in hard midday sun, from the dark pillars inside to the open rooftop. The sound is Ne-Yo's "Closer" ("…under a spell I can't break, I just can't stop").
Why it suits AI: it needs two people, a lot of energy, a lot of takes and a whole car park. Two photos cover all of that, and the twenty cuts are exactly the kind of editing nobody wants to redo by hand.
The 23 Seconds, Shot by Shot
| Time | What happens |
|---|---|
| 0:00 | Over photo 1's shoulder down a corridor of pillars; photo 2 is a tiny figure far away, running in |
| 1.3s | Photo 1's close-up: the sunglasses come off, a deadpan stare |
| 2.8s | Photo 2 sprints at the camera |
| 3.4s | A ground-level shot of photo 2 bounding past |
| 4.7–7.2s | Photo 1 spins and crouches as photo 2 closes in; a hand covers the lens |
| 7.2s | The face-off: photo 2 throws an arm up, crouches and springs up |
| 9.0s | Photo 2's close-up, palm to the lens |
| 9.7s | Photo 2 shoves photo 1 back by the neck and chest |
| 12.0s | Photo 2's first close-up, singing |
| 15.1s | Photo 2's second close-up: "I just can't stop" |
| 16.1s | Photo 1's close-up, grinning |
| 17.1s | The wide rooftop battle; photo 2 lunges across and out, photo 1 is left alone |
| 21.5s | Photo 1 dances alone by the brick wall |
| 22.8s | Back to the opening shot |
Your Look
- Your own faces and outfits. Nobody gets the original's black sweater, track jacket or baggy trousers.
- Your own roles. Photo 1 always waits, photo 2 always runs in and sings.
- Your own sunglasses. Wear them in your photo and photo 1 takes them off in the second shot.
Put whoever has the better sunglasses moment in slot 1 and the better singer in slot 2. Full-length photos give the best outfits, because the wide shots show you head to toe.
The Fastest Way
Closer (Dance)
Two photos, all twenty shots, the lip-synced close-ups and the rooftop battle, with the original sound.
Two photos, 250 credits (about $12.50), no prompt. It renders on Seedance 2.5 at 480p, vertical 9:16, 23 seconds.
Build It Yourself
1. Turn the clip into a depth map. Run it through a depth-estimation model (Depth Anything V2 works). Smooth the runner's curly hair into a plain dome and blur the glasses, so neither shows up on you. Keep the song inside the depth video: the model needs to hear the lyric to time the close-ups.
2. Make one storyboard sheet of six moments. The sunglasses coming off, the sprint, the shove on the rooftop, the singing close-up, the grin and the wide battle, generated from both photos in the car park, with a depth-map sheet of the original poses as a layout guide.
A photorealistic still from a handheld phone video on a sunny concrete car park rooftop, apartment towers behind a low parapet: the person from reference image 2 puts one hand on the neck of the person from reference image 1 and the other on their chest, pushing them back in a staged dance-battle move, while the person from reference image 1 leans back playing along, each wearing exactly the clothes from their own photo. Hard midday sun. No text.
3. Drive it with the depth map and the song. Feed the sheet, both photos and the depth video to a model that copies motion from a reference video (we recommend Seedance 2.5). Then write what it can't read from a depth map: a timed list of all twenty shots, who is who in each, exactly when photo 2 sings and that photo 1 never does. Render it and lay the original audio back on at the end.
Common Mistakes
Not saying who starts where in the wide shots. In the far rooftop shot the two figures are tiny, and the model can swap them, or briefly turn one into a copy of the other where they cross. Write down who starts on which side and who exits where.
Leaving the curls in the depth map. Curly hair reads clearly in a depth map and ends up on whoever plays that part. Smooth it into a dome first.
Both mouths singing. Without saying that only photo 2 sings, and only in the close-ups, the model has both of you mouthing the lyric through the whole battle.
Writing the shove as a fight. Describe it as staged choreography, or the model reaches for real violence.
FAQ
How many photos? Exactly two: photo 1 waits, photo 2 runs in and sings.
How long is it? 23 seconds, vertical 9:16.
What does it cost? 250 credits, about $12.50.
Does it work with pets? No, this one is built for people.
Related Reading
- How to Make the "What's Up In A Minute" AI Video: the swiss side quest trio dance, cut on the beat
- How to Make the 'Smooth Criminal Remix' AI Dance Video: a couple dance in the place from your photo
- How to Make the '58th Floor' AI Video: you and a friend rapping at the COLORS mic
- How to Make the 'Look Alive' AI Video: you and a friend in the Drake and BlocBoy JB video
- Viral AI Video Trends 2026: every trend we've turned into a template this year
