VellmHow to make the Closer dance AI video with a friend
By Vellm ·
The Closer dance AI trend is a comedy clip filmed on a tree-lined park path. One friend stands waiting in sunglasses, slowly takes them off, and the other comes sprinting down the path toward the camera. What follows is a fast, slightly unhinged face-off to Ne-Yo's "Closer", ending with the two of them dancing side by side.
To make the Closer dance video with your own faces, upload two photos to the Closer Dance template, one of each person. Neither of you has to run anywhere: you get 15 seconds of vertical video (9:16) for 50 credits.
What happens, beat by beat
It cuts fast, so this is the rough shape rather than every frame:
| Time | What you see |
|---|---|
| 0–1.25 s | Over the waiting person's shoulder, a tiny figure appears at the far end of the path. |
| 1.25–2.75 s | Cut to the waiting person, who takes off the sunglasses and stares into the lens. |
| 2.75–4.75 s | The sprinter runs at the camera, a tree trunk wipes across the frame, and we see them side-on, mid-stride. |
| 4.75–7 s | The shot opens down the path. The waiting person reacts and throws up a hand to stop them. |
| 7–12 s | The face-off: quick cuts, both of them in each other's faces, grabbing at each other's arms. |
| 12–14 s | A close-up of the sprinter, then one wide look at the face-off. |
| 14–15 s | The two of them dance side by side on the path. |
The sunglasses moment is two seconds long and it carries the whole joke. Everything after it is the payoff.
Which photo goes where
Photo 1 is the waiting person: the one in sunglasses who takes them off. Photo 2 is the sprinting person, who runs in from the far end of the path. If one of you has the better deadpan stare, put them in Photo 1.
Both photos are required, and each needs one adult, with the face and upper body clear. A few things help:
- No sunglasses in the photo itself. The template adds the pair the waiting person takes off. Your photo needs to show the eyes.
- Large, front-facing faces. A face that fills a good part of the frame gives the template more to work with than a small face in a wide shot.
- Two separate photos, one person in each, not a group shot. The template needs to know who plays which role.
- People who have agreed to be in it. Your own photo, or a friend who said yes.
Your photos give the faces and the clothes. The park path, the trees and the moves come from the original clip. Other people who pass through the original clip are generated as new fictional adults, so they will not look like anyone you know.
Generating it
Put the waiting person in Photo 1 and the sprinter in Photo 2, then generate. No account is needed for a one-time purchase, and the video usually takes about 10–15 minutes. When it arrives, watch the face-off closely: that is where the two of you are nearest each other, and the place to check that each person kept their own look, especially if you look alike.
Adding the song
Without the music the face-off is just two people flailing, so you will want "Closer" back. Download the silent version for posting and add the song inside TikTok or Reels, if it is available for your use there. The other download keeps the original clip's sound, for personal viewing only.
What it costs
Two friends, 50 credits: $6.99 for the 50-credit pack that covers it. Rotating different friends through the sprinting role? On the 600-credit pack each video works out to $4.58. There is no subscription; pricing lists every pack.
For photo advice that applies to every template, see how to choose a reference photo.