Full length, standing
Shoes included: the outfit in the photo is what you will be running and dancing in.
Send in one photo and the Sydney station clip belongs to you: the run down the steps, yellow doors sliding shut, knees on the concrete, a heap on the platform, then a full dance on the concourse. The shaky camera and the beats land where they always did. 15 seconds, sound included.
Short answer: One photo is enough: DanceRemake drops you into the viral Missed Train Dance, the run down the station stairs, yellow doors closing on you, knees on the concrete, a heap on the platform and then a dance in the middle of the station. You get 15 seconds, 9:16, carrying the clip's audio unless you ask for silence. Sign in with Google, send one sharp photo and press Remake this. One video uses 10 credits ($9.99 bought on its own); nothing is generated for free.
Your photo1 needed
A single person with the face in clear view, no sunglasses or strong filters. Head-to-toe photos give the best result, because you sprint and dance in the clothes you wear in it.
Soundtrack
Your video plays with the original clip's audio.
What you getFixed
Sign in with Google at the next step.
Failed renders are refunded automatically. Photos are deleted after 30 days.
What changes
More examples

The trend
TikTok creator @zacjeremiah filmed 15 seconds inside a Sydney train station and posted them on May 5, 2026. He runs down the stairs, the yellow doors close in front of him, he drops to his knees and lands flat on the platform, then gets back up and dances it out on the concourse, with commuters and orange-vested station staff looking on. Over 10 million people have watched it. From Oct 18, an editing template started dropping strangers into those same 15 seconds, and from Oct 22 the AI versions took off across Britain and the United States. People look it up as Missed Train Dance AI, as the missed train trend, or simply as the Missed Train Dance.
What you get is those 15 seconds filmed again with the person from your photo in the lead. Same staircase, same doors, same concourse, same hand-held camera, same beat, only your face and your clothes. Leave the trend sound on and the clip's own audio sits under your video; switch it off for a silent file and pick the track in TikTok when you post. DanceRemake has no tie to the person who filmed the clip or to the musicians behind its song, and the file you download is a fresh AI render rather than his footage.
What makes people repost it is the hard cut in mood: one second he is face down in defeat, two seconds later he is dancing as if the train never left. Everyone has watched a door close on them, and since the camera never cuts away, a remake has to keep his timing, his sprint and his knee slide intact. Here that is exactly what gets filmed again, with somebody new in the lead.
Photo tips
Shoes included: the outfit in the photo is what you will be running and dancing in.
A single person looking at the camera, no shades, no heavy filters. The face reads best when you turn back to the lens to dance.
The dance is timed to the clip's own audio, so only drop it if you plan to add the track inside TikTok.
How-to
A single person with the face easy to read. Head to toe is best, because whatever you are wearing in the photo is what you run and dance in.
Leave Trend sound on to keep the clip's own audio. Pick No song and you get a silent video, ready for the track you add in TikTok.
Roughly five minutes later a 15-second MP4 shows up in My videos, watermark-free. Mark it as AI-generated wherever you post it.
Before you pay
Yes. Each clip is rendered fresh on paid AI models, so there is no free tier. One video takes 10 credits ($9.99 for a single one, less per video in bigger packs), and the MP4 you download carries no watermark.
DanceRemake is a fan-made tool and nothing more, with no tie to whoever filmed the clip or to the musicians behind its song. Your download is a brand-new AI render that restages the scene around you.
Anyone can watch the examples on this page and the clip they came from, no account needed. Google sign-in only shows up when you go to make your own.
Refunds: when a render fails its credits return to your balance by themselves, and a credit pack bought by mistake and left unused can be refunded inside 14 days.
FAQ
Yes, as long as Trend sound stays on: those 15 seconds of audio from the clip ride under your video. Choose No song instead and you get silence, which lets you pick the track inside TikTok or Reels. Reused audio gets muted on some platforms, so that route is the safer one.
Not at all. A photo is the whole job. The AI carries you through every part of it: the stairs, your knees hitting the floor, the sprawl, then the dancing, landing on the same beats as the man in the original.
It can, as long as the face is sharp. Full length is stronger: with only your head and shoulders, the AI has to invent your pants and shoes for the running and dancing.
Not here. Sprinting, dropping onto your knees and dancing all need human legs. Pets do have their own version of our AI elf dance.
Add one sharp photo here, keep Trend sound on or switch to No song, then press Remake this. Google sign-in comes next, you pay for one video, and five minutes or so later the 15-second file is in My videos.
No. Every clip is a real render with a real bill behind it, so there is no free tier and no sample with a watermark on it. A video costs 10 credits, and a failed render puts those credits straight back.
Five minutes or thereabouts, a bit longer when the queue fills up. Feel free to close the page; your video will be sitting in My videos.
Your face, hair, build and outfit carry over, for women and men alike, which the two examples on this page show. The biggest factor is a bright, sharp full-length photo; small things such as logos or jewelry may still come out soft.
No. What you download is a clean MP4 for TikTok, Reels or Shorts. Tick the AI-generated box if the app offers one.
Yes, a genuine phone recording made in a Sydney train station and posted to TikTok on May 5, 2026. Everything in the AI wave that followed, including what you make here, is newly generated video with a different person in it.
Who it's for
The one who misses trains, buses and flights. Send it straight to the group chat.
Put yourself on the platform for a morning everyone recognizes.
Cast the birthday person, or the colleague whose leaving party is this week.
Up next
Your crew sings the chorus into one phone in Seoul.
You in the cowboy hat, strutting through the quinceañera.
Dozens dance in formation. You do not move at all.
Two of you, side by side, on the red-walled routine.
Squat low, grip handlebars nobody can see, outrun the hover-bike cop.
A new dance blew up today?
We add this week's dances as soon as they take off.