AI Dance Video Examples

Six relay dance videos from our own test runs. For each one you can see the two photos that went in and the raw 13.7-second file that came out.

How these examples were made

Every example below was made by the same generator you use on our AI Dance Video Generator page, with the 13-second setting and the same dance instructions. Each one started from two photos of two different people. Four people took part in these tests, and all four agreed to have their photos and videos shown here.

The videos are the generator’s raw output, exactly as it returned them: 13.7 seconds long, vertical 9:16, with the AI-generated soundtrack it added. Nothing was cut, re-ordered, slowed down or retouched. When you make a video yourself, the file you download is trimmed to exactly 13 or 15 seconds and delivered at 720×1280.

We ran 8 tests in total and show 6 of them here. The two we left out looked very much like examples 4 and 6: one person on screen for under two seconds. Showing them as well would only repeat the same point.

Six relay dance examples, photos and results

Input: photos A and B
Input photo A: a woman in a cream top and pleated skirt standing in front of a beige wallInput photo B: a man in a black cardigan, white T-shirt and grey trousers on a tree-lined path
Output: 13.7-second unedited original
Example 1 · photos A + BThe person from photo A dances in front of the wall for about 8 seconds. Then the video cuts to the person from photo B on the tree-lined path for the last 5.6 seconds.
Input: photos C and D
Input photo C: a man in a black T-shirt and white trousers on a road beside a grass fieldInput photo D: a woman in a navy hoodie and navy trousers on a lawn with palm trees
Output: 13.7-second unedited original
Example 2 · photos C + DThe person from photo C dances for about 9 seconds, starting on the road; partway through, the background shifts to a dance studio. Then the person from photo D takes over on the lawn for the last 4.6 seconds.
Input: photos A and B
Input photo A: a woman in a cream top and pleated skirt standing in front of a beige wallInput photo B: a man in a black cardigan, white T-shirt and grey trousers on a tree-lined path
Output: 13.7-second unedited original
Example 3 · photos A + BThe person from photo A dances by the wall for about 10.7 seconds without a cut. The person from photo B gets the last 3 seconds on the path.
Input: photos A and B
Input photo A: a woman in a cream top and pleated skirt standing in front of a beige wallInput photo B: a man in a black cardigan, white T-shirt and grey trousers on a tree-lined path
Output: 13.7-second unedited original
Example 4 · photos A + BThe person from photo A is on screen only for the first 1.8 seconds. The person from photo B dances on the path for the remaining 11.8 seconds.
Input: photos A and B
Input photo A: a woman in a cream top and pleated skirt standing in front of a beige wallInput photo B: a man in a black cardigan, white T-shirt and grey trousers on a tree-lined path
Output: 13.7-second unedited original
Example 5 · photos A + BThe person from photo A starts by the wall, then at 1.9 seconds the video cuts to an indoor studio where her clothes differ from the photo, and they differ again after 6 seconds. The person from photo B dances on the path for the last 3.8 seconds.
Input: photos C and D
Input photo C: a man in a black T-shirt and white trousers on a road beside a grass fieldInput photo D: a woman in a navy hoodie and navy trousers on a lawn with palm trees
Output: 13.7-second unedited original
Example 6 · photos C + DThe person from photo C dances for about 11.5 seconds, mostly in a dance studio after a cut at 1.4 seconds. The person from photo D appears only for the last 2.2 seconds, on the lawn.

What the examples show

  • Both people appeared every time. In all six videos, the person from each photo is on camera and dancing, one after the other.
  • The split between the two people is never the same. The first person’s turn ran from 1.8 seconds (example 4) to about 11.5 seconds (example 6). You can’t set this, and neither length option gives you control over it.
  • Backgrounds and clothes can drift. In examples 2, 5 and 6 the first person ends up dancing in a studio that wasn’t in the photo, and in example 5 her clothes differ from the photo twice. The second person stayed closer to the original photo in all six.
  • The format never changed. Every clip is vertical 9:16 and the full length requested.

In these six, the person from the first photo danced first. We still don’t promise the order, because the AI makes that call, but it gives you a sense of what usually happens.

Why we show raw files

AI video is unpredictable, and polished highlight reels hide that. We would rather you see the real spread of results before you spend anything: some videos give both people a fair share, and some give one person a short cameo. If an even split matters a lot for what you want to post, these examples tell you what to expect. The details of what is fixed and what varies are on our how it works page.

Make your own

Upload two photos on the home page and try a free 5-second preview, or see pricing for full 13 or 15 second videos without a watermark.