Gangnam Style AI
Video Generator from Photos
Turn 1–3 photos into a Gangnam Style-inspired AI dance video with Gamgamstyle AI. Start with a solo photo or explore an experimental multi-person request.
Trio mode is experimental. Character consistency, choreography and final results may vary. Explore the trio beta.
Generate AI Dance Video →Upload 1–3 Photos
Choose clear photos to get started, with one person per reference. Three-photo submissions use the experimental trio workflow; placement, likeness and movement can vary.
JPG, PNG or WEBP · up to 10 MB per photo · one person per photo · clear body and face · use images you have permission to upload
The length options are requests to the video model. Final duration, dimensions and dance accuracy can vary. This tool does not automatically crop every image to 9:16.
Watch real solo and duo outputs
Both clips below were generated by this tool on October 22, 2026 with the built-in dance prompt. The solo clip’s one-second warm-up opening was trimmed so it starts on the dance; the duo clip needed no trimming. These examples show solo and duo output, not a tested 19-second trio. Nothing else was edited.
Want to inspect this duo before making a request? See three real frames and the framing changes in our sample review. It explains what this short clip shows and what it cannot prove about a trio.
How to make a Gangnam Style AI video from 1–3 photos
Choose clear photos
Use up to three photos — one person or original character each, facing the camera, with the torso and legs visible where possible. Use images you have permission to upload.
Request a clip
Choose a duration supported by your confirmed access. Longer requests require approved access: solo up to 15 seconds, duo or trio up to a 19-second target. Rendering time and results vary.
Review and save the MP4
Check movement and likeness before posting. Download links last one hour; check the saved video again to refresh the link while the task remains available.
Built around one workflow
1–3 photos in
Upload one to three JPG, PNG or WEBP photos — start with a solo or explore an experimental duo or trio request. The browser resizes them before sending — no filming, no template, no timeline editing.
Dance prompt included
The preset horse-riding choreography prompt is built in. Add optional character notes to steer the look, outfit and setting.
Private by default
Photos, task records and finished videos stay in a private Cloudflare R2 bucket and are served through short-lived signed links.
Resume, don’t redo
The latest task is saved in your browser. Re-check it or refresh an expired link without generating a new video.
MP4 you can review
Download the finished MP4 and inspect motion, likeness and framing before posting. Crop or edit it anywhere afterwards.
Four languages
The same workflow runs in English, Español, Русский and Türkçe from one site, with shared guides in each language.
Continue the same video
The latest task is saved in this browser. If checking times out or the connection drops, return to the generator and choose “Check saved video / refresh link”. It checks that task without generating a new video. Keep this browser’s site data; “Start a new video” replaces the saved task.
Get the shape you want
Output dimensions tend to follow the input photo. Our verified test used a landscape photo and returned a 1236 × 744 landscape clip; upload a vertical photo when you want a vertical clip. Dimensions are still not guaranteed — inspect the MP4 before posting.
Three ways to make your next dance
Guides, prompts and comparisons
“Gamgamstyle” is our name for AI character videos inspired by the horse-riding dance from PSY’s Gangnam Style. These guides cover the term, the workflow and the prompt behind the generator.
Character ideas (English):
Compare other workflows (English):
Before you hit generate
Is Gamgamstyle AI free?
No free trial is currently offered. Contact Billycui50@gmail.com for package pricing, available access and possible discounts. Online checkout and automatic credit purchases are not active.
Can three people dance together from three photos?
The generator accepts one to three JPG, PNG or WEBP photos, one person or character per photo. Two or three references are sent to a group video model. Three-person likeness, choreography and identity stability can vary; faces can merge or change. Read the Trio tutorial before uploading.
Which lengths can the generator request?
The form offers 5, 10, 15 or 19 seconds. Single-photo generation caps at 15 seconds; a 19-second request needs two or three photos. The preview defaults to 5 seconds and longer requests need approved access. A 19-second request is a duration target, not an exact recreation of a reference video.
Is 480p or 720p output automatic?
The current models generate at 720p or above. A 480p package needs a separate export arranged by email; this website does not yet automatically downsample or deliver paid package versions. Output dimensions and framing can vary. A 720p source does not reduce generation cost when later exported at 480p.
What if I close the page or the connection drops?
The latest task ID is saved in the same browser. Return and select “Check saved video / refresh link” to check that task without submitting another generation. Keep the site data; starting a new video replaces the saved task.
Does the video include Gangnam Style music?
The generator does not supply the original song or a music licence. Add audio available for your account and use case in TikTok or an editor, and follow the platform’s AI disclosure rules. Song availability differs by region and account. Read the TikTok music guide.