Real generated output · 2 photos

Watch a two-photo Gangnam Style AI duo

Two reference photos become two dancers in one short clip. Watch the original published output below.

Real duo output · two reference photos · 5.042 seconds · no trimming or other edits
Published duration
5.042 seconds
MP4 dimensions
1280 × 720
Editing
Original output · untrimmed
Clip first published

This is the unedited 5.042-second output from a five-second duo request. It contains two characters, not three, and it does not demonstrate a 19-second result.

Behind this clip

How this example was made

The duo used two reference photos, one character per photo, with the generator’s built-in horse-riding dance prompt. The result is a 1280 × 720 landscape MP4. No trimming or additional visual editing was applied to the published clip.

The finished file was checked on : 5.041667 seconds, 1280 × 720 pixels, 24 frames per second, one video stream and no audio stream. The values above describe this public MP4.

Inspect the three extracted duo frames

Watch with intent

What to check in the output

  1. Body count and motion

    Watch both dancers through every movement. The yellow jacket and green hoodie help you distinguish the two performers, but the input photos are not published here to verify likeness.

  2. Changing composition

    At 0.5 seconds, both dancers and their shoes are visible. By 2.5 and 4.5 seconds, the framing is tighter and the lower legs leave the image. Check this before a vertical crop.

  3. Sound and limits

    This MP4 has no audio stream. Add permitted music separately. A short two-person sample cannot establish three-person consistency or the quality of a longer request.