Gangnam Studio · Updated October 26, 2026
How to Make a Gangnam Style AI Video
How to make a Gangnam Style AI video: start with a reference
The current three-person selfie trend puts a trio together in one scene. It is different from a tutorial for the original 2012 horse-riding choreography. Watch the reference clips on our homepage to understand the framing and movement; they were not generated by this site.
1. Prepare three portraits
Choose one clear photo of each consenting adult. Use a single visible face, even lighting and a simple background. Avoid sunglasses, heavy face filters, cropped heads, motion blur and other people in the frame. A clear input helps preparation, but output identity and quality have not yet been validated.
2. Assign Left, Center and Right
Confirm photo rights before uploading. Add one JPG, PNG or WebP portrait to each position, up to 10MB per file. Check the thumbnails and use Replace or Remove if you chose the wrong photo. All three positions are required to save a request.
3. Save and return to My Videos
Choose Save request · no charge. Your request is stored on the server and appears in My Videos with the three private photos and Waiting to start status. Refreshing the page preserves it when you use the same browser session.
What Waiting to start means
Generation is not open. Your request has not been submitted to a provider, no payment has been collected, and there is no estimated completion time. It will not automatically turn into a paid generation later. Photos and requests expire after seven days; delete them earlier from My Videos if you prefer.
The planned output
The intended format is a 15-second 720p MP4, with the three one-time packs shown on the homepage. Purchases, generation and downloads will require a separate validated release. The reference examples do not establish that the service can deliver identical results today.
Build your trio · My Videos · Privacy · Terms