How to Make a Gangnam Style AI Video from Your Photos

Start with one clearly visible subject in each image. For people, an evenly lit face looking toward the camera is a useful starting point. Avoid sunglasses, masks, strong beauty filters and distant faces. A group photograph creates ambiguity about which person belongs in a role, even if everyone in the photo is recognizable to you.

Use JPEG, PNG or WebP files up to 10 MB each. The prepared upload also needs both dimensions between 301 and 5,999 pixels and a width-to-height ratio greater than 0.4 and less than 2.5. If an image is rejected, follow the displayed size or format message. The photo guide explains suitable examples.

Open the Gangnam Style studio. Gangnam Style trio uses three separate references. Upload the first solo subject to Middle, the second to Right and the third to Left. Left and right describe the viewer’s view of the scene. Check each preview before continuing; the box order is an assignment system, not just a place to collect files.

Choose Three of you for one selfie used across all three roles, or Pet trio for three individual pets. Changing modes clears the current photo selection, so decide your cast first. The one-selfie guide covers the repeated-subject option in more detail.

Framing is fixed at 16:9 landscape to retain the reference composition. Choose 480p for a lower-resolution output or 720p for a higher-resolution file. The choices currently cost 150 and 300 Gangnam Style AI credits respectively. Check the quote beside the controls before submitting; selecting quality does not itself start a paid generation.

The clip follows an approximately 15-second template. Resolution describes output size, not a guarantee of stronger likeness or better anatomy. Choosing 720p cannot reliably repair an unclear reference or a mistaken role assignment. Resolve those issues in the photo previews first.

Check your current account or guest wallet balance and confirm that you own the photos or have the necessary permission. If the interface asks you to sign in, follow that prompt. Then submit once. Uploading, preparing the video and adding the soundtrack are separate processing steps; follow the studio status until the completed file is ready.

You can check an active job in My videos. If a network request fails, use the explicit retry control and read the message instead of repeatedly starting new versions. A confirmed failed generation returns its reserved credits. The cost and troubleshooting guide explains those cases.

Play the completed video and inspect the opening group shot, each solo and the ending. Check that the subjects remain recognizable through transitions. A good still frame is not enough to judge a moving result. New generations include the reference clip’s soundtrack; use the player controls and device volume to listen.

Download the MP4 and play that local file as well. This distinguishes a playback setting from a problem in the exported file. Save a copy you can find again before preparing a post. Existing older videos are not automatically given a new soundtrack simply because you reopen their history.

The landscape file can be uploaded as a regular YouTube video. YouTube’s Shorts guidance calls for square or vertical files up to three minutes, so a short duration alone does not make this 16:9 output a Short. If you reframe it elsewhere, review the side performers carefully rather than cutting them out unintentionally.

Follow the destination’s music and AI-disclosure requirements. YouTube requires disclosure for realistic, meaningfully altered or AI-generated depictions. A generated performance of real people can fall within that guidance. Use a caption that describes the clip as an AI creation rather than footage of something your friends actually did.