3
JPEG / PNG / WebP
Gangnam Style AI Dance
Upload three photos for the Gangnam Style trio selfie: each friend gets a turn at the camera. Create a 15-second landscape video with the reference clip’s audio.
More video models
Explore other ways to animate your photos.
Gangnam Style AI Dance
Three friends. One camera. Everyone gets a turn.
Gangnam Style AI Dance uses your three photos and the built-in reference video with MiniMax H3 Max. Photos guide appearance; the video guides positions, solo shots, movement and camera timing. The original soundtrack is added afterward.
Upload your photos768p · 15s · 16:9Describe your scene
768p · 15s
16:9 · MiniMax H3 Max
Shown before creation
The estimate includes output duration, reference-video processing and applicable image-input charges.
Gangnam Style AI Dance
Choose the cast by the opening positions
Photo 1 is the center person, photo 2 the person on the left and photo 3 the person on the right in the opening shot.
Follow the selfie handoffs
The trio opens together.
Give each face room to read
Close selfie framing highlights all three friends while showing your chosen background. Describe a beach, café or another location and its lighting. Avoid sunglasses, heavy filters and group photos; check faces and hair during camera transitions in the result.
Three friends. One camera. Everyone gets a turn.
Positions refer to the opening group shot. Upload one clear person in each slot; the camera later moves between them.
Choose the cast by the opening positions
Photo 1 is the center person, photo 2 the person on the left and photo 3 the person on the right in the opening shot. Use separate, clear portraits with visible eyes and hair. These roles stay assigned as the camera moves; they do not change with each person’s position on screen.
Follow the selfie handoffs
The trio opens together. The center friend leads, the camera swings to the opening right friend, then to the opening left friend, before returning to the center. Small shoulder movements, face-level gestures and the hands-on-head moment matter more than large footwork. The close framing is part of the effect.
Give each face room to read
Close selfie framing highlights all three friends while showing your chosen background. Describe a beach, café or another location and its lighting. Avoid sunglasses, heavy filters and group photos; check faces and hair during camera transitions in the result.
How to create your video
01
Upload your photos
Use one clear JPEG, PNG or WebP photo in each position.
02
Describe your scene
Describe a location and lighting if you want to change the scene. Leave blank to follow the reference video’s setting; cast assignments and dance direction are built in.
03
Generate and review
Sign in, check the credits and create. Review faces, hands and movement before downloading.
Questions
Is this the classic horse-riding dance?
No. This page follows the three-friend street-selfie performance shown in the reference, with rotating close-ups and small gestures. It is not a full-body recreation of the classic choreography.
Do I need to write a prompt or upload a video?
Upload three photos and optionally describe the scene. The built-in video reference is attached automatically; you do not need to upload a video. The original soundtrack is added after generation.
What video settings are used?
MiniMax H3 Max generates a 15-second landscape video using the 768p tier and a 16:9 request.
Does the video include sound?
The audio from the reference clip shown here is added to the finished video at its original speed and pitch. It is not a newly generated performance of the song.
Will every face and movement match exactly?
Photos guide identity and the reference video guides movement and framing. Faces, hands, solo framing and timing may still vary. Exact beat and lip synchronization are not guaranteed; review the full result.
How are credits calculated?
The estimate includes output duration, reference-video processing and applicable image-input charges.
Make the next clip yours.
Upload three photos for the Gangnam Style trio selfie: each friend gets a turn at the camera. Create a 15-second landscape video with the reference clip’s audio.
