← Back to Blog

How to Make the Gangnam Style AI Trio Video

By Summrs Team••4 min read

Not the Horse Dance

If your feed is full of three friends singing Gangnam Style into a phone on a city street at night, that is not a 2012 throwback. On October 5, 2026, TikTok creator @emiliyyam posted a clip captioned "gangnam style in gangnam." Three friends on a Gangnam street — Korean flags on the lampposts, office towers lit up behind them — passing one selfie camera back and forth while PSY's song plays. Group shot, solo close-up for each person with whip pans between them, and a palm pushed at the lens. It hit 19 million views and 3.1 million likes.

By October 12, AI remakes started appearing. By October 19, the trend had split into character trios, friend-group versions, and CapCut templates each pulling millions of views. The format people call "Open Gangnam Style" or "Oppa Gangnam Style" is this — the trio selfie, not the horse-riding dance.

The Gangnam Style Trio template on Summrs puts your three people into the night selfie. Three photos in, one video out.

Try AI Photo Editing, Color Grading & Video Generation

Summrs analyzes each photo and applies professional edits automatically—color grading, object insertion, restoration, viral video generation and more. Describe what you want in plain English, and see results in seconds.

Try for Free →

How to Make the Gangnam Style AI Video

  1. Open the Gangnam Style Trio template on Summrs.
  2. Upload the centre dancer — this person holds the camera, opens and closes the video, and gets the most screen time.
  3. Upload the second person — they get their own solo close-up with a whip pan.
  4. Upload the third person — they get the last solo before the camera returns to the group.
  5. Pick 720p or 480p and generate.

You get a landscape 16:9 clip with the song. All three roles are required — each person needs their own photo.

Try AI Photo Editing, Color Grading & Video Generation

Summrs analyzes each photo and applies professional edits automatically—color grading, object insertion, restoration, viral video generation and more. Describe what you want in plain English, and see results in seconds.

Try for Free →

Why Character Trios Took Over

The format is built for three, and the internet already has favorite trios. Within two weeks of the original, the most-viewed AI remakes were:

  • Harry, Ron, and Hermione at Hogwarts in the snow — 1.3 million views
  • Gandalf, Frodo, and Aragorn from Lord of the Rings — 1.1 million views
  • Tobey, Andrew, and Tom as the three Spider-Men — 676,000 views
  • Turkish actors in a version that hit 3.6 million views

The format works because the joke is in the casting. Three people who would never film a night selfie together, singing Gangnam Style on a Korean street. The less likely the trio, the better the video performs.

Your version works the same way. The funniest ones are not three friends who actually went out — they are three people who have no business being in the same selfie.

Try AI Photo Editing, Color Grading & Video Generation

Summrs analyzes each photo and applies professional edits automatically—color grading, object insertion, restoration, viral video generation and more. Describe what you want in plain English, and see results in seconds.

Try for Free →

Who Goes Where

Photo assignments map directly to the original video:

  • Photo 1 — Centre / phone holder. Opens and closes the video. Gets two solo moments. The most screen time by far.
  • Photo 2 — Second friend. Gets a solo close-up with a hair flip or hand gesture.
  • Photo 3 — Third friend. Gets the last solo in front of a glass building, hands in their hair.

Put the person the video is about in Photo 1. For a birthday, that is the birthday person. For a group chat roast, that is whoever lost the bet. For a character trio, the most recognizable face goes centre.

Try AI Photo Editing, Color Grading & Video Generation

Summrs analyzes each photo and applies professional edits automatically—color grading, object insertion, restoration, viral video generation and more. Describe what you want in plain English, and see results in seconds.

Try for Free →

Photos That Work

Three people in close-up selfie lighting means the AI needs strong identity signals from each photo:

  • One person per photo — no group shots. The model needs to assign one identity per role.
  • Face visible and well lit — most of the video is close-ups. Shadows, sunglasses, or hair covering the face will weaken identity.
  • Different outfits help — three people in similar dark clothes blur together during the group shots. Contrasting colors give the model more to work with.
  • Night-out clothes match the vibe — everyone keeps their outfit from their photo. A leather jacket, a white turtleneck, a beanie — whatever you are wearing is what you wear on the Gangnam street.

Output is 16:9 landscape with the song already on the video. Post to TikTok, Reels, or Shorts.

For more group AI templates — Charlie Wilson Tiny Desk 5-person, BTS SWIM Dance 5-person, ATEEZ BAD 3-person, and more AI face swap video templates.

FAQ

Is this the horse dance? No. The 2012 music video horse-riding dance is a different format. This is the October 2026 trio selfie trend — three friends passing a phone camera on a night street.

What does "Open Gangnam Style" mean? "Open" is how many people hear "Oppa" (오빠), the first word of the famous line "오빠 강남스타일." They are the same trend.

How many photos do I need? Three. Centre, second friend, and third friend. All three are required.

Does the song play? Yes. PSY's "Gangnam Style" plays through the entire video.

Can I use it for characters or pets? Any clear photo with a visible face works — real people, illustrated characters, or pets. The character trio versions are some of the most-viewed remakes of this trend.

Related

Ready to Transform Your Workflow?

Edit photos, color grade entire shoots, and generate AI videos—all in one platform. Just describe what you want in plain English, and Summrs handles the technical work.

Try 10 Photos Free →