The Gangnam Style AI trend, explained
Three people on a terrace at night, a song-and-dance number and a few quick camera moves. Here is what the clip is, how its beats work and why people keep remaking it.
Updated September 25, 2026
Where it comes from
The name comes from "Gangnam Style", PSY's 2012 hit. More than a decade later its official music video is still being watched: Korean press reported that it passed 6 billion YouTube views on 17 July 2026.
The clip the trend is built on comes from an older film: three men doing a song-and-dance number on a terrace at night, with potted plants, lamp posts and a lit glass building behind them. The center singer opens in close-up, the camera whips over to each side for a solo, and it ends back on the center close-up.
The trend took off on TikTok as people recast the same trio with other faces. Popular versions put famous football players in a stadium, a racing team on a rooftop at night, or characters from fantasy and wizard films into the clip. The remakes are short, usually around 15 to 19 seconds, and most people make theirs with a CapCut template or an AI generator.
Why the format works
The clip is short and the structure is simple, so any three people swapped in still read as the same video at a glance.
- Three spots, three moments. Everyone gets a solo of their own, which makes it a natural format for groups of friends, teams and families.
- Big close-ups. The center singer fills the frame at the start and the end, so a new face is the first and last thing people see.
- Quick camera moves. The whip pans between the singers keep the clip moving and make each swap land as a small reveal.
- A name people know. "Gangnam Style" has been familiar since 2012, so the trend is easy to search for and recognize.
How the clip works, beat by beat
The generator on this site makes a 15-second version that follows these beats.
0:00–0:04
The center person sings into the lens in close-up while the other two lean in behind.
0:04–0:08
A whip pan to the left person, who sings and gestures alone.
0:08–0:12
The right person sings and dances alone in front of a glass building.
0:12–0:15
Back to the center person in close-up, singing and waving an open hand at the lens.
How people make it now
There are three ways, and they differ mostly in how closely they follow the beats.
- One-click generator. Upload three photos, press Generate video and get the same beats with your people in them. Open the generator
- CapCut template. Tap Try this template on a video and add your photos or Live Photos; the result depends on the template. CapCut steps
- Your own prompt. Describe the three people, the night terrace and the beats in Kling, Veo or Sora; results vary a lot. Copy the prompts
The step-by-step guide covers photo tips and a full comparison.
Ready to put your three in the clip?
Upload three photos, press Generate video, and get a 15-second MP4 with the original audio in about 10 minutes.