Gangnam Style AIGuide
How to make your Gangnam Style AI video
Updated October 29, 2026
Start with three photos and a fixed dance template. You do not need to write a prompt, choose choreography or edit a timeline.
1. Choose the three people and photos
Ask each person for permission to appear in an AI dance video. Everyone must be 18 or older, and you must have the rights to every photo. Use one clear, well-lit portrait per person with an unobstructed face.
2. Fill the three performer roles
Add a PNG, JPEG or WebP portrait for the Left, Center and Right Dancers, shown from the viewer’s perspective. Check every preview and replace any wrong or unclear image. All three photos are required before generation.
3. Sign in and choose credits
Each render costs 100 credits. Sign in with Google or email, then choose a one-time pack if you need credits. There is no free render or subscription. Review the three photos again before starting.
4. Wait for the result
Progress appears in stages while the video is created. There is no guaranteed completion time. You may leave and return to the generator or account history. If the job fails, its 100 credits return automatically.
5. Download and label the video
The result is a 15-second, vertical 9:16 MP4 at 720p with the same fixed soundtrack heard in the demo and no watermark. Say that it is AI-generated when posting and check the sharing rules of the platform you use.
Questions before you start
Can I use one group photo?
No. The template needs three separate photos, each showing one consenting adult, assigned to the Left, Center or Right Dancer position from the viewer’s perspective.
Can I change a photo after the video starts rendering?
Review the previews before generating. A running job uses the photos submitted for that job. Changing the inputs does not revise a video already in progress.
What if the result has a problem?
Failed jobs automatically return 100 credits. For a completed but clearly broken result, follow the refund policy and include the video link. AI likeness and motion can vary.