Rumpelstiltskin AI: do the viral tiptoe dance with a friend
Two photos, two roles: one sneaks in, shushes the camera and tiptoe-dances on the hay in pointy curled shoes; the other watches, trying not to laugh. About 12 seconds, landscape, with sound.
No prompt
Two photos, two roles
~12 s · 16:9 · with sound
×
JPG, PNG or WebP · under 20 MB · one clear, front-facing face0/2 photos
Use everyday photos of yourself or friends (with their consent). Celebrity photos and film or MV stills won't pass review.
Only upload photos of people who agreed to be featured.
Three beats, twelve seconds
The moves, close-ups and cuts follow the viral original. Your photos decide who is in them.
0–4 s
The shush
Photo 1 creeps in between the hay bales, puts a finger to their lips and grins. Photo 2 covers their mouth, then pulls a face.
4–8 s
The tiptoe dance
Arms spread wide, knees bent, Photo 1 tiptoes side to side across the hay with a coat tail swinging, while Photo 2 watches from the foreground.
7–12 s
The pointy shoes
Low close-ups of pointy, curled-toe shoes skipping on the hay, cut against a smug arms-out grin. That's the Rumpelstiltskin bit.
Sample
The same barn motion track. Swap the people and the clip looks brand new.
Sample
Made from two AI sample photos
Two fictional AI people take the tiptoe dance and the reaction, with the hair and outfits from their photos.
The two photos used
Tiptoe dancerOnlookerThe result
Caption ideas
The ways people use this meme most. Tap one to copy it, then paste it when you post.
Make yours in three steps
01
Add two photos
Photo 1 does the shush and the tiptoe dance, photo 2 reacts. One person per photo, face clearly visible.
02
Pick quality and generate
480p costs 15 credits, 720p 30 credits, 1080p 50 credits. Photos are reviewed first, renders usually take a few minutes, and failures are refunded automatically.
03
Download or share
The clip comes with its soundtrack. Download the MP4, or share it straight to TikTok, Reels or Shorts from your phone.
Get a better result from the first upload
Front-facing, sharp, evenly lit
One person per photo, nothing covering the face. Both roles get big close-ups, so a clear upper-body shot works best.
Hair follows the photo
Each photo's hairstyle is read and kept in the video — same length, curl and color — so pick a photo with hair you like.
So does the outfit
Whatever someone wears in the photo, they wear in the video. The dancer always gets the pointy curled shoes on top.
Most dramatic friend in photo 1
Photo 1 carries the shush, the grin and the whole dance; photo 2 gets the reaction shots. Cast accordingly.
A viral clip set in a hay-filled barn: one person sneaks up, shushes a friend, then tiptoe-dances like Rumpelstiltskin in pointy curled shoes while the friend tries not to laugh. This template recasts both roles with you and a friend, keeping the original moves, close-ups and cuts.
As far as meme trackers can tell, it started as an AI-generated clip on Instagram in late August 2026: a fairy-tale gnome tiptoe-dancing with a sly grin, styled like a 1980s fantasy film. Paired with a rap track in late September, it took off as the go-to reaction for sneaky moments — or for being far too proud of something small.
It's most often paired with Riff Raff's 2014 track "Tip Toe Wing In My Jawwdinz" — the tiptoe steps land right on its beat. Our clips come with the original video's music added automatically when you play or download them.
No. The original gnome clip was itself AI-generated in the style of an old fantasy film, so there's no movie or actor behind it. Here, the stars are you and your friend.
15 credits per 480p video, 30 credits per 720p video and 50 credits per 1080p video. If a photo isn't approved or the render fails, the credits are refunded automatically.
One person per photo, a clear front-facing face and even light. Use photos of yourself or friends (with their consent); celebrity photos and film stills won't pass review.
Yes. Put your photo in both slots and you play the dancer and the onlooker.
Yes. The render itself is silent; the original track is added in your browser when you play or download it, synced to the picture.
Photo review usually takes a few seconds to a minute or two, then rendering takes a few minutes. You don't need to keep the tab open — finished clips are saved to Your renders.
