GANGNAM STYLE AI TRIO · NIGHT SELFIE · 1–3 PHOTOS
Gangnam Style AI Trio Generator
Gangnam Style AI trio turns 1–3 selfies into the October 2026 three-friend night selfie set to PSY’s “Gangnam Style”. Choose the phone holder, left friend and right friend for a 15-second video with music.
1–3 photos · 15 s · 16:9 · Music included · From 30 credits · Failed videos refunded
Photo 1 · Center / phone holder
One person per photo. An empty place repeats Photo 1.
Photo 2 · Left friend (optional)
One person per photo. An empty place repeats Photo 1.
Photo 3 · Right friend (optional)
One person per photo. An empty place repeats Photo 1.
300–6000 px on each side, up to 30 MB. Use one clear, front-facing person per photo, with hair and clothing visible. Crop a group photo into three separate portraits before uploading; faces are not separated automatically.
480p · 15 s MP4 · 16:9 · the trend’s sound on the video
This video costs 30 credits. Charged when you start, refunded automatically if it fails.
30 credits is about $0.83 at the Studio monthly rate. The $5 Trial pack (24 credits) is not enough for this 15 s video.
Gangnam Style AI examples
The 15-second night selfie: group shot, the phone holder’s close-up, two solos and the palm to the lens. 16:9, with sound.
What is the Gangnam Style AI trio?
Gangnam Style AI trio is a video remake of the 2026 night selfie trend: three friends share a handheld camera, sing along to PSY’s “Gangnam Style” and take turns in close-up. This generator casts your photos into a fixed night-terrace version and exports a 15-second 16:9 MP4 with music. It costs 30 to 864 credits, from about $0.83 at the Studio monthly rate.
Before you upload
| What you upload | 1–3 separate portraits, one person per photo. Photo 1 fills any empty places; crop a group photo first. |
|---|---|
| Length | 15 seconds, the whole selfie |
| Frame | Horizontal 16:9, so all three faces fit in the group shot |
| Resolution | 480p (default) or 720p; 1080p on Seedance 2.0 Pro and Seedance 2.5; MiniMax H3 at 720p |
| Sound | The trend’s soundtrack is included. Generated mouth movements can drift from the audio. |
| Models | Seedance 2.0 Mini (default), Seedance 2.0 Fast, Seedance 2.0 Pro, Seedance 2.5, MiniMax H3 |
| Cost per video | 30–864 credits, from about $0.83 at the Studio monthly rate; one to three photos cost the same |
| Free? | No free generated preview; sign-up credits and the $5 Trial pack do not cover one 15 s video |
| Watermark | Videos made before any purchase carry a moving tryh3.com mark; none after a purchase |
| If it fails | A rejected photo or failed generation refunds the credits. A completed result you choose to redo is a new paid generation. |
Which photo gets each Gangnam Style trio solo?
Positions mean the viewer’s left and right in the opening group shot. The close-up order is Photo 1 → Photo 2 → Photo 3 → Photo 1. Leaving a slot empty repeats Photo 1; it does not add a random person.
| You upload | Phone holder (group shot, about 1.6–4.5 s, final close-up) | Left friend (solo about 4.8–8.2 s) | Right friend (solo 8.45–11.9 s) |
|---|---|---|---|
| 1 photo | You | You again | You a third time |
| 2 photos | Photo 1 | Photo 2 | Photo 1 again |
| 3 photos | Photo 1 | Photo 2 | Photo 3 |
Have one group photo? Save three crops, with one person in each, then place them in the three slots. This tool does not detect and assign everyone from one group upload.
What a Gangnam Style AI trio video costs
The model and resolution set the price; one, two or three photos cost the same. Dollar equivalents below use Studio monthly ($49.90 for 1,800 credits). They are credit values, not single-video checkout prices.
| Model · length | 480p | 720p | 1080p |
|---|---|---|---|
| Seedance 2.0 Mini · 15 s | 30 ($0.83) | 65 ($1.80) | — |
| Seedance 2.0 Fast · 15 s | 65 ($1.80) | 139 ($3.85) | — |
| MiniMax H3 · 15 s | — | 155 ($4.30) | — |
| Seedance 2.0 Pro · 15 s | 109 ($3.02) | 233 ($6.46) | 584 ($16.19) |
| Seedance 2.5 · 15 s | 156 ($4.32) | 351 ($9.73) | 864 ($23.95) |
Lowest charge: Mini at 480p, 30 credits (about $0.83 at the Studio monthly rate). The $5 Trial pack gives 24 credits and is not enough by itself. The smallest purchase covering a video is $29.90: a 700-credit Boost pack or the 800-credit Creator monthly plan.
Higher resolution adds pixels, but does not guarantee a better likeness or lip-sync. Mini and Fast offer 480p/720p; Pro and Seedance 2.5 add 1080p; MiniMax H3 offers 720p.
Five Mini 480p videos use 150 credits (about $4.16 at the Studio monthly rate); ten use 300 credits (about $8.32). A new attempt after a completed video costs the displayed price again.
Free preview, payment, watermark and download
You can watch the examples free. Generating your own trio requires an account and enough credits; there is no short generated preview to unlock afterward.
| Question | Without a purchase | After any purchase |
|---|---|---|
| Account and upload | Free account; uploading a photo does not start a charged generation | The same account and upload process |
| Full 15 s video | Sign-up credits do not cover one (30 credits minimum) | 30–864 credits, shown before you generate |
| $5 Trial pack | 24 credits, 6 short of the minimum by itself | Additional eligible credits can cover the balance |
| Watermark | A video generated before any purchase carries a moving tryh3.com mark | New videos generated after a purchase have no site watermark |
| Download | An existing completed result can be downloaded as MP4 | MP4 from the result or Creations; no separate download charge |
| Rejected photo or failed generation | Charged credits are refunded | Charged credits are refunded |
| Free generated preview | Not offered; the examples can be played free | Generation produces the full 15 s video |
| Redo a completed result | A new generation requires its own credits | Charged again at the selected model and resolution |
Gangnam Style AI trio video examples
Four versions of the 15-second night-terrace routine: a group selfie, the phone holder’s close-up, a left-friend solo, a right-friend solo, then the phone holder again. Each includes music. Faces, clothing and gestures can vary between generations.
Make your Gangnam Style AI trio in four steps
Use the generator above in a phone or desktop browser. No filming, software installation or prompt writing is required.
- 01
Pick who plays who
Choose the opening center person for Photo 1, the viewer’s left friend for Photo 2 and the viewer’s right friend for Photo 3. Separate selfies are fine; you do not need to be together.
- 02
Add the photos
Upload one person per photo. Crop a group photo into separate portraits first. One selfie fills all three places; with two selfies, Photo 1 also fills the right-hand place.
- 03
Choose model and resolution
Pick a model and resolution. Mini at 480p costs 30 credits; the button shows the full charge before generation. An extra face does not add to the price.
- 04
Generate and download
Sign in and generate when you have enough credits. When the video finishes, download the MP4 from the result or Creations. Check each close-up before sharing.
Gangnam Style AI trio template: the 15-second shot order
The template uses a night terrace, a handheld selfie camera and quick pans between the three people. Times below are approximate; generated gestures and mouth movements can vary.
| Time | What happens |
|---|---|
| 0–1.6 s | Group selfie at arm’s length: the phone holder in the middle, the left friend with a hand behind the head, the right friend leaning in, then raising a hand beside the phone holder. |
| 1.6–4.5 s | Close-up of the phone holder singing the chorus straight into the lens. |
| 4.5–4.8 s | A fast whip pan. |
| 4.8–8.2 s | The left friend alone: hand on chest, a shoulder sway, a loosely open hand reaching toward the lens, then raised-hand gestures beside the face and shoulders. |
| 8.2–8.45 s | Another whip pan. |
| 8.45–11.9 s | The right friend alone: lifts a loosely open hand outward and steps closer, moves a curled hand near the chest, raises an open hand beside the head, then turns side-on and steps back. |
| 11.9–12.1 s | A last quick pan. |
| 12.1–15 s | The phone holder again: an open palm close to the lens, lowered out of view around 13.7–14 s, followed by head turns and the final mouth movements. |
Motion blur during the quick pans is part of the style. Check the stationary close-ups for face quality instead of judging a single blurred transition frame.
Does the Gangnam Style AI video include music?
Yes. The downloaded video includes the trend’s soundtrack with PSY’s “Gangnam Style” chorus. Audio is placed on the finished clip; the model’s mouth movements may not line up perfectly in every shot.
If all mouth movements are offset by the same amount, you can adjust the audio timing in an editor. If only one person’s lips drift, moving the whole track may throw the other solos off; a new generation may still differ.
Check the music permissions for your intended use before posting, especially for brand content or ads. Purchasing generation credits does not grant a licence to the song.
The night-selfie trend behind this template
The October 2026 trend follows @emiliyyam’s “gangnam style in gangnam” selfie: three friends share a camera in Seoul and each sings a short solo to PSY’s 2012 song. AI remakes cast new people into that format. This page uses the night-terrace variation shown above; it does not make the horse-riding choreography from PSY’s original music video.
Trend and song references: The 2026 Gangnam selfie · PSY’s official music video
Gangnam Style AI trio face swap: fixing mixed-up faces
Start by checking the photo order and the input portraits. A clean photo helps, but AI can still change a face, hand or expression. A higher price is not a guarantee of a more accurate result.
| Problem | What to check | Fix |
|---|---|---|
| The face does not look like you | Face too small, side-on, obscured or heavily filtered | Use a sharp, front-facing portrait with even light. Keep the full face and hair visible. |
| Two friends swap faces between shots | Wrong upload slot, several people in one photo, or similar-looking portraits | Photo 1 is center, Photo 2 is viewer’s left, Photo 3 is viewer’s right. Use one distinct portrait per slot; check all three previews before generating. |
| Only the face changed, not the hair or clothes | A very tight crop may omit the hairstyle and clothing | Use a chest-up or waist-up portrait showing the hair and outfit you want. |
| Lips drift off the chorus | Mouth animation differs from the fixed soundtrack | Check whether every solo has the same offset. A simple offset can be edited; inconsistent mouth animation may require a new paid attempt. |
| Faces blur when the camera moves | Quick camera pans, soft input photos or limited output detail | Pause after the pan. If the face remains soft, use a sharper portrait and consider 720p or 1080p for more pixels. |
| A friend is cut off when posting vertically | Cropping the 16:9 video to 9:16 | Post it as 16:9, or place it on a 9:16 canvas with a blurred background instead of cropping |
Photos that work and photos that do not
| Works well | Avoid |
|---|---|
| One person, looking at the camera | Group photos with several faces |
| Chest-up or waist-up, outfit visible | Tight face crops or full-length shots from far away |
| Even daylight or soft indoor light | Dark, backlit or heavily filtered photos |
| Your own face, or friends who agreed | Celebrities, public figures or anyone who did not agree |
Do I need a Gangnam Style AI video prompt?
No. Select the portraits, model and resolution; the trio routine is already set. You do not need to describe the camera moves or write a face assignment prompt. The upload slots determine which person plays each role.
This is a fixed 15-second night-terrace template. It has no custom background, choreography or prompt editor. Use the same role order when replacing a photo, then check the new price before starting another generation.
Gangnam Style AI trio FAQ
Which photo goes in the middle, and who sings first?
Photo 1 is the phone holder in the middle of the opening selfie and the first close-up. Photo 2 is the friend on the viewer’s left, followed by Photo 3 on the right. The phone holder returns for the ending: 1 → 2 → 3 → 1.
Can one selfie make all three people look like me?
Yes. Upload it as Photo 1 and leave Photos 2 and 3 empty. Photo 1 fills all three roles. With two uploaded portraits, Photo 1 plays the phone holder and right friend, while Photo 2 plays the left friend.
Can I upload one group photo for the Gangnam Style AI trio?
Crop the group photo into three separate portraits first, with one person per image, then upload them into the intended slots. This tool does not automatically separate or assign several faces from one group photo. Three separate selfies also work.
Is there a free full video or a free generated preview?
No. Watching the examples is free, but making your own full video starts at 30 credits. Sign-up credits and the $5 Trial pack of 24 credits do not cover one by themselves. There is no short generated preview; the smallest purchase covering a video is $29.90.
When do I pay, and does fixing a finished video cost again?
Uploading does not start a charged generation. The Generate button shows the charge: 30 credits (about $0.83 at the Studio monthly rate) to 864 credits, depending on model and resolution. Starting generation spends credits. Failed generation or rejected photos refund them; redoing a completed result is a new paid attempt.
Does the download include the song, and will the lips match?
The MP4 includes the trend’s Gangnam Style soundtrack. Mouth movements are AI-generated, so exact lip-sync is not guaranteed. Check each solo before sharing; a single audio shift only helps if the offset is consistent across the clip.
Why does the left friend get the phone holder’s face?
Check that Photo 2 contains only the left friend and Photo 1 only the phone holder. Avoid group uploads, obscured faces and heavily filtered portraits. Even with correct inputs, the model can mix identities between shots; a new attempt may help but is charged again.
How do I post the landscape trio as a vertical Reel?
Download the 16:9 MP4 and place the whole frame on a 9:16 canvas in your editor, using bars or a background to fill the space. Cropping the group shot to fill a vertical frame can cut out the two friends. This generator exports landscape only.
Make your Gangnam Style AI trio
Choose the phone holder and two friends, check the price, then make the 15-second night selfie with music.
Upload your photos