How to make a Gangnam Style AI trio video
Prepare three separate portraits, assign the solo order, and make the Gangnam Style AI trio trend.
By Groove Editorial · Updated October 26, 2026
Start with the right three people
The trio template requires three separate photos, one consenting adult per image. Photo one is the middle person and first solo; photo two is the right person and second solo; photo three is the left person and final solo. Choose portraits where each face is clear and nobody is hidden behind another person. A straight-on composition gives the template a more useful starting point than a dramatic camera angle. Avoid very small faces, dark sunglasses, heavy motion blur, and screenshots with interface elements. Use an original photograph when you have one. Upload only pictures you own or have permission to use, and obtain permission from everyone shown.
Prepare the photograph
Crop each portrait separately to show one person clearly. Do not crop through a face or include a second person in the same image. Natural lighting and a simple background can make subjects easier to distinguish. A sharp photo does not guarantee a perfect result: generated movement may change facial details, clothing, hands, or the background. Think of the output as a playful transformation, not a faithful recording of a real performance.
Select the trio template
Open the generator, choose the trio option, and upload all three prepared portraits together. Review the preview before submitting. If the wrong image appears, replace it before starting a generation. The template supplies the dance direction, so you do not need to write a complicated choreography prompt. The service requests a 15-second video at your selected 480p or 720p quality from the configured video provider. The generated output is silent. Add audio separately only when you have the necessary rights.
Wait for the actual result
Video generation is an asynchronous task. Keep the page open while the job is processing and follow the status displayed by the generator. Completion time depends on provider availability and workload; there is no guaranteed generation time. A loading animation is a status indicator, not proof that a finished video exists. If a job fails, read the error before trying again. Repeated submissions can create separate provider jobs, so avoid submitting the same request multiple times while it is still processing.
Check before sharing
Watch the entire clip and inspect all three faces, their movement, and the edges of the frame. Look for identity changes, unusual limbs, flashing details, or unintended background content. Download a result only when a finished video is available. Ask the other people in the photo before posting publicly, and describe the clip as AI-generated when sharing it. Do not present it as evidence that the people actually danced or endorsed a product. If the photograph is difficult for the model, try a clearer source image rather than expecting repeated runs to repair it automatically.