Make the Oppa Gangnam Style AI trend your own with a friend or your whole trio. Recast the duo street dance with deadpan poses, a position switch, and the iconic horse-riding move. Or take over the nighttime terrace in a three-person remix with solo close-ups and a group finale. Pair your images with the matching reference video and copy-ready prompts to put your cast in the spotlight.
Choose two images for the duo or three for the trio, using a compatible multi-image reference-video workflow. Keep the reference format: vertical 9:16 for the street duo or landscape 16:9 for the terrace trio.
Give your Gangnam Style AI video a cast everyone recognizes. Assign one clear image to each performer and specify consistent faces, hairstyles, and outfits through position changes, solo close-ups, and the final group shot.
Turn two friends into the Gangnam Style duo street meme. One stands in front with folded arms while the other dances behind, then they exchange positions before the shared horse-riding sequence. Follow your reference for the full comic payoff.
Create a Gangnam Style Trio AI video with three friends on a nighttime terrace. Give the center, left, and right performers their own camera moments, then bring everyone back together. Follow the reference's lip-sync, gestures, and individual expressions.
Match the duo or trio reference format to keep every performer in view. Use 9:16 for the vertical street duo and 16:9 for the wide terrace trio. For a vertical trio edit, reframe each shot so the side performers keep their moment.
Prepare two images for a duo or three for a trio, with one person clearly visible in each. Choose sharp faces and visible outfit details; full-body views help with the duo's footwork. Use your own images or those of adults who agreed to the edit.
Add the matching Gangnam Style reference video in a compatible multi-image reference-video mode. For the duo, assign Image 1 to the person initially in front and Image 2 to the dancer behind. For the trio, assign Image 1 to center, Image 2 to screen-left, and Image 3 to screen-right in the opening shot.
Generate your Oppa Gangnam Style AI dance video and check every performer's face and outfit across the full sequence. Keep the reference's duration and aspect ratio within the selected model's limits. Download your favorite take, align your music, and reframe for your chosen social platform.
Two identity images + the vertical street reference
Cast two friends in the cool-versus-chaos street dance. Prompt: “Use Image 1 as adult Person A, initially in front, and Image 2 as adult Person B, initially dancing behind, in Video 1. Preserve each face, hairstyle, skin tone, outfit, shoes, and proportions throughout. A begins with folded arms and a deadpan expression while B dances energetically. Follow the reference as they exchange positions, perform the side-by-side horse-riding steps, and approach the camera. Exchange positions and performance roles, never identities. Match the reference's exact actions, timing, expressions, gaze, camera movement, framing, cuts, brick-wall street, and lighting. Keep footwork balanced, hands coherent, and landings grounded. Retain vertical 9:16 and the reference duration within the model's limits. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No face blending, extra people, invented choreography, sliding feet, outfit flicker, text, or watermark.”
Three identity images + the landscape terrace reference
Give all three friends their own close-up in the terrace remix. Prompt: “Use Image 1 as adult Person A at center, Image 2 as adult Person B at screen-left, and Image 3 as adult Person C at screen-right in Video 1's opening group shot. These labels follow the same people throughout, including solo shots. Preserve each face, hairstyle, skin tone, outfit, accessories, and proportions. Follow the exact reference sequence: group, center close-up, left solo, right solo, center return, then the group reunion. Match each performer's gestures, mouth movements, expressions, gaze, and timing. Preserve all original camera moves, transitions, cuts, terrace details, skyline, doorway, and nighttime lighting. Keep landscape 16:9 and the reference duration within the model's limits. Do not add horse-riding steps absent from the clip. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No face swaps, identity blending, extra people, altered outfits, invented shots, text, or watermark.”
One identity image + one selected duo performer
Join the duo by replacing just the dancer who starts in the background. Prompt: “Use Image 1 as the identity reference for one adult in Video 1's Gangnam Style street duo. Replace only the performer who starts dancing in the background. Track that same person through the position switch, side-by-side dance, and close-ups. Keep the other performer's original identity, outfit, and actions unchanged. Preserve Image 1's face, hairstyle, skin tone, outfit, footwear, and proportions. Follow the selected performer's exact choreography, poses, expressions, mouth movements, and timing. Retain the original street, brick wall, lighting, framing, camera movement, cuts, duration, and 9:16 format within the model's limits. Keep exactly two performers wherever both appear in the reference. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No identity transfer to the partner, face blending, extra steps, distorted hands, sliding feet, outfit flicker, text, or watermark.”
Three identity images + a trimmed 5–10 second trio reference
Test the trio's solo-to-group identity consistency with a short reference segment. Prompt: “Use Image 1 as the adult center performer, Image 2 as the adult screen-left performer, and Image 3 as the adult screen-right performer, defined by the full trio reference's opening group shot. Use a trimmed 5–10 second segment of that reference as Video 1. Keep these identity labels fixed even if the segment opens on a solo close-up. Reproduce only the shots present in the trimmed clip, in their original order, with matching gestures, mouth movements, expressions, gaze, camera moves, cuts, terrace background, and nighttime lighting. Preserve each person's face, hairstyle, outfit, accessories, and proportions across every reappearance. Keep 16:9 and match the segment's duration with a compatible model setting. SAME IDENTITY. NO DUPLICATE. NO NEW SCENES. No added group shot, horse-riding dance, automatic loop, face merging, extra people, wardrobe flicker, text, or watermark.”
The Oppa Gangnam Style AI trend uses personal images to recast a reference performance set to PSY's song. The duo street version features deadpan poses, a position switch, and horse-riding steps. The trio terrace version moves between three performers' close-ups and a group finale. Choose the matching reference before assigning your cast.
Use two separate identity images for the duo or three for the trio, with one clear face per image. For the trio, assign Image 1 to center, Image 2 to screen-left, and Image 3 to screen-right in the opening shot. Choose visible outfit details and unobstructed faces. Use images of yourself or consenting adults.
An image and a prompt can guide a new horse-riding dance animation. To follow the duo's position switch or the trio's solo-shot sequence, use the corresponding reference video in a compatible workflow. The full-cast prompts use two or three identity images; the single-performer replacement uses one. Supported inputs and duration depend on the selected model.
Assign every image to a named performer before generating. For the duo, positions and performance roles can change while identities stay fixed. For the trio, center, left, and right refer to the opening group shot and remain the same identity labels during solo shots. Check returning faces, outfits, accessories, and the final group composition.
The prompts guide the visual performance; naming “Oppa Gangnam Style” does not automatically include PSY's song. Audio depends on the selected model and settings. Preview the result, then align an authorized track in your editor or use the publishing platform's music library where available. Keep the music aligned with your reference sequence.
Create Images and Videos
