What happens in the Raindance clip, shot by shot?
Each version copies a viral AI reel, and your two photos replace its two characters. The pier, the sea, the camera moves and the timing come from the reel. Both versions follow the same 15 seconds, the opening of the trend:
| Seconds | What happens |
|---|---|
| 0 to 4 | Photo 1 sits alone at the end of a wooden pier and looks out to sea. The camera moves around behind them, and photo 2 appears far down the pier. |
| 4 to 7 | Photo 2 sings straight to the camera, with the sea behind. |
| 7 to 10 | The two sit side by side on the sand, and photo 2 sings to photo 1. |
| 10 to 15 | Back by the water, photo 2 reaches out to the camera, then sings in a close-up. |
Human or Pet? Press "Create your own" and pick one, or press the button under its reel on this page.
| Human (default) | Pet | |
|---|---|---|
| The reel it copies | @unchained_off and KevkoYt, 8 October 2026 | @yikes_301, two AI-made cats, 25 September 2026 |
| Photos | Photo 1 listens, photo 2 sings | Pet 1 listens, pet 2 sings |
| Frame | Vertical 9:16, ready for TikTok and Reels | Wide 16:9, a crop of the 4:3 cat reel |
| Length | 15 s | 15 s |
| Sound | The reel's song, the same 15 seconds; you can turn it off | The same |
Both versions use MiniMax H3 Max Recast. It puts your photos into the reel itself, so the shots, the camera and the singer's lip movement stay the reel's, and the faces, bodies and clothes come from your photos. The singer's face fills the frame for most of the clip and the lips have to follow the song, so both versions use this model rather than a cheaper one that makes a new scene. The exact price shows on the Generate button before you spend anything. Every run is a new take, so results differ.
