Step into the role of the mysterious figure in velvet: give a sly "shh," perform the unsettling tiptoe routine through the straw, and watch a blonde princess react in disbelief. Upload a friend's picture to put them in the role of the astonished onlooker.
Rumpelstiltskin AI features a peculiar character in a glossy velvet suit who gestures for silence and tiptoes through a nighttime hay barn while a shocked princess looks on. Though it resembles a 1987 movie scene, this was created entirely by AI in 2026. Now you can become that dancer—all you need is a single sharp photo.
Also called Rumpelstiltskin 1987 · Tip Toeing in My Jordans · Rumpelstiltskin meme · Rumpelstiltskin dance · Tip Toe Wing In My Jawwdinz · AI Rumpelstiltskin Tip Toein' In My Jordans
Example
What your Rumpelstiltskin Dance video looks like
Made on Smew AI from an AI-made photo with the prompts on this page, not a real person. Your result will differ from run to run.
Made with Smew AI
Made from one AI-made photo with the The dancer, face clear version, Face only option (228 credits). Nobody in it is a real person.
The source AI footage from @stroinaya: the character peeks, hushes, then tiptoes across hay as the princess gasps. Original clip mode places your face onto this figure.
How to create the Rumpelstiltskin Dance on Smew AI
Everything happens in your browser. No editing software, no prompt writing, no watermark.
Upload your photo
Tap "Create your own" and upload the dancer's photo with a visible face. Capture with your camera or select from your files.
Tap Generate
The cost appears on the button. Seedance 2.5 inserts your photos into the Rumpelstiltskin trend footage and recreates it frame by frame for 15 seconds, with the soundtrack from the trend example. No prompt required.
Download or post
Your video arrives within several minutes as a 16:9 MP4. Save it, post from your device, or retrieve it later from your Library.
What makes a good photo?
Choose a clear, well-lit face photo looking toward the camera or slightly away. The system preserves your face, hair and facial hair while dressing you in velvet, so a head shot works fine.
Face completely visible: no shades, hats, masks or strong filters.
Good lighting, ideally natural. Dark or fuzzy selfies force the model to guess features, producing inaccurate results.
One person only per photo.
In Original clip, your face replaces the dancer's while he keeps the velvet suit. The princess remains the woman from the source footage.
Choose Whole body in Original clip to substitute the entire person, pet or character from your image, keeping their original outfit or appearance. Full-body shots work best here.
In Original clip, 2 people, upload the dancer first, princess second. The second photo supplies her face and hair while she keeps the white nightgown.
For You and a friend, upload the dancer first, observer second. The second image becomes the nightgown-wearing observer who gasps.
Use your own face or a consenting friend's. Never celebrities or strangers: we don't create those, despite their popularity in viral swaps.
Where the Rumpelstiltskin Dance trend began
This footage is AI-generated, not from any actual movie. AI creator Tatyana Tereshchenko (@stroinaya) shared it on Instagram on 27 August 2026, crediting @neuroferma and @anna_glinskaya, with a Russian caption describing a fantasy film "without a camera and actors". A small figure in velvet dances for a young woman in a hay-filled barn where straw transforms to gold.
TikTok added the soundtrack a month afterward. On 29 September @6aimbot9 paired it with Riff Raff's 2014 song "Tip Toe Wing In My Jawwdinz", reaching 926,000 plays by 8 October. An X repost from @itsjusttimbro on 1 October hit 13.6 million views. Many reposts captioned it "Rumpelstiltskin (1987)", sparking searches: US Google queries for "rumpelstiltskin 1987" peaked 5 October. The actual 1987 Rumpelstiltskin is a Cannon musical starring Billy Barty and Amy Irving, without this scene. By then, people were inserting themselves, friends and TV characters as the dancer.
AI creator Tatyana Tereshchenko (@stroinaya) shares the source footage on Instagram as an example of AI-generated fantasy filmmaking. Instagram
TikTok creator @6aimbot9 sets the footage to Riff Raff's "Tip Toe Wing In My Jawwdinz"; by 8 October the post reaches 926,000 plays. TikTok
An X repost from @itsjusttimbro spreads the footage globally, surpassing 13.6 million views by 8 October. X
Green-screen variations emerge: @deckcake places the dancer before a refrigerator captioned "creeping downstairs at 2am for a lil treat", surpassing 1.1 million TikTok plays. TikTok
Face-swap versions dominate: @ns7films' version featuring two TV characters reaches 400,000 TikTok plays by 8 October, while US Google searches for "rumpelstiltskin" rise tenfold above baseline. TikTok
Know Your Meme publishes the entry "AI Rumpelstiltskin 'Tip Toein' In My Jordans'"; NBC News reports on the meme the following day. Know Your Meme
Copy and paste
Rumpelstiltskin Dance prompts
You don't need these: the single-tap process handles them automatically, along with the trend footage. [Image 1] and [Image 2] represent your photos, [Video 1] represents the trend footage being replicated.
Original clip
1 photo: the dancer, face clearSeedance 2.515 s · 720p · 16:9228 credits
Video prompt
Recreate the reference video [Video 1] shot for shot: the same 1970s fairy-tale film look, 35mm grain, faded warm color, the same hay barn at night, the same blonde woman in the white nightgown, the same camera angles, cuts, timing and dance moves. The ONLY change: the small dancer with the wild curly blond hair is now the person in [Image 1].
The dancer = the person in [Image 1], in EVERY shot where the head is visible: copy the exact face from [Image 1] - face shape, eye color and eye shape, eyebrows, nose, mouth, skin tone, age, hair color, hair length and hairstyle, and any beard, moustache, glasses or freckles exactly as in the photo. The same real person in every frame, from the first shot to the last, so their friends instantly recognize them. Never morph, never drift back to the original actor's face, never add or remove facial hair. Keep the original dancer's body size and costume: the dark green-black velvet tuxedo jacket with long coat tails, white shirt, black bow tie, dark trousers and pointed black shoes.
Shots, matching the reference:
1. Close-up of the blonde woman: she turns around in the hay, gasps and covers her mouth with her hand.
2. Wide: the dancer peeks out of the dark doorway behind her; the back of her blonde head in the foreground.
3. Wide: the woman sits in the hay in her white nightgown and looks around.
4. Medium shot: the dancer stands in the hay, holds a finger to the lips "shh" with a sly grin; the back of her blonde head in the foreground.
5. Close-up of the dancer's face with the finger on the lips, sly look - clearly the person from [Image 1].
6. Extreme close-up of the woman's eyes, her hand on her mouth.
7-8. Wide: the dancer tiptoes toward the camera and does the bouncy tiptoe dance across the hay, knees bent, arms spread wide, coat tails flapping.
9. Medium close-up: arms spread wide, grinning, bobbing to the rhythm.
10. Low shot of ONLY the legs and pointed black shoes tiptoeing through the straw - legs, trousers and shoes only, NO face, NO head anywhere in this shot.
11. Close-up of the woman, both hands pressed to her mouth.
12-13. The dancer raises one finger and grins: first wide, then a medium close-up.
14. Wide: the woman sits among gold coins in the hay while the dancer runs past her to the right.
Anatomy rules: one head only, on the dancer's shoulders. A face never appears on knees, legs, hands, clothes or the hay. The blonde woman keeps her own face and her long blonde hair in every shot, also in the last wide shot. No text, no logos, no modern objects.
1 photo: your photo, face clearWan 3.010 s · 720p · 9:16128 credits
Video prompt
A 10-second vertical 9:16 fairy-tale film, one continuous shot, 1970s 35mm look, heavy grain, warm amber hay and teal shadows. Night inside an old wooden barn full of golden hay. The dancer is the person in [Image 1]: keep their face, skin tone, eye color, hair, hairstyle and any facial hair, a little messy, so friends recognize them. They wear a shiny dark emerald-black crushed-velvet tuxedo jacket with long tails, a white shirt, a black bow tie, dark trousers and pointed black shoes with curled toes. An invented young woman with long golden-blonde hair, in a white cotton nightgown, watches from the side. She never gets the dancer face. Exactly two adults.
0-2 s: the dancer peeks from behind a haystack and holds one finger to the lips, face large and sharp.
2-4 s: the woman gasps, a hand over her mouth.
4-7 s: the dancer tiptoes through the loose straw toward the camera, knees bent, shoes crunching in the hay.
7-8.5 s: they stop, one finger raised, the curled shoes sunk in the straw.
8.5-10 s: a close shot, they break into a broad mischievous grin, the face large and sharp.
Audio: Tip Toe Wing In My Jawwdinz by Riff Raff, plus straw rustle. No speech. No text, no logos, no watermark. Photorealistic, not CGI, not a cartoon.
2 photos: you, the dancer + your friend, the princessSeedance 2.515 s · 720p · 16:9228 credits
Video prompt
Recreate the reference video [Video 1] shot for shot: the same hay barn, cuts and tiptoe dance. The ONLY changes: the little man is now the person in [Image 1], and the woman in the white nightgown is now the person in [Image 2].
The dancer keeps the velvet tuxedo and the woman keeps the white nightgown. Copy each face, hair and facial hair from its own photo. Never swap them and never drift back to the original faces. Exactly two people. No text, no logos, no watermark.
2 photos: you, the dancer + your friend, who watchesWan 3.010 s · 720p · 9:16128 credits
Video prompt
A 10-second vertical 9:16 fairy-tale film, one continuous shot, 1970s 35mm look, a hay barn at night, warm golden light on the straw. The dancer is the person in [Image 1]: their own face and hair, in a dark velvet tuxedo with long tails, a white shirt and a black bow tie. The observer is the person in [Image 2]: their own face and hair, in a white nightgown. Never swap the two faces. Exactly two adults.
0-2 s: the dancer peeks from behind a haystack and holds one finger to the lips, face large and sharp.
2-4 s: the observer gasps, a hand over the mouth.
4-7 s: the dancer tiptoes through the loose straw toward the camera, knees bent.
7-8.5 s: they stop, one finger raised, shoes sunk in the straw.
8.5-10 s: a close shot, the dancer breaks into a broad grin, the face large and sharp, while the observer watches.
Audio: Tip Toe Wing In My Jawwdinz by Riff Raff, plus straw rustle. No speech. No text, no logos, no watermark.
You pay in credits. The Generate button shows this price before you spend anything. These are the options on this page.
Option
Model
What you get
Credits
Original clip: Face only
Seedance 2.5
15 s · 720p · 16:9
228
Original clip: Whole body
Seedance 2.5
15 s · 720p · 16:9
228
You dance
Wan 3.0
10 s · 720p · 9:16
128
Original clip, 2 people
Seedance 2.5
15 s · 720p · 16:9
228
You and a friend
Wan 3.0
10 s · 720p · 9:16
128
One-time packs and plans are on the pricing page. A failed video gives its credits back.
Does the Rumpelstiltskin 1987 dance come from an actual film?
No. This tiptoe footage was entirely AI-generated in 2026. It convincingly mimics the aesthetics of vintage European fairy-tale cinema: gentle focus, warm vintage color grading, film grain, an unsettling close grin and a heroine with 1960s-style eyeliner.
The mix-up stems from captions. Numerous reposts labeled it "Rumpelstiltskin (1987)", and an actual movie exists with that title: a 1987 Cannon musical featuring Billy Barty as Rumpelstiltskin and Amy Irving as the miller's daughter. That movie contains no velvet outfit or tiptoe sequence. Some guessed it originated from vintage Russian cinema; while the creator's caption is Russian, the footage itself is recent.
As a meme format, the character represents foolish dancing, sneaky behavior, or bizarre responses to situations. Successful captions included: "The type of dreams I be having after taking dollar store melatonin", "creeping downstairs at 2am for a lil treat" and "when my wife asks how many beers I've had but I don't remember".
Which Rumpelstiltskin format should you choose?
Four versions are available.
Version
What you get
Length and frame
Sound
Original clip (default)
One photo. Seedance 2.5 swaps your face onto the little man throughout the actual AI clip, in all shots and close-ups. Choose Whole body to replace him entirely with your full photo (a person, pet or character). The princess, barn, and all cuts and movements remain identical.
15 s, wide 16:9 matching the original
The soundtrack from the trend example
Original clip, 2 people
Two photos, little man first. Uses the same actual clip, and applies your friend's face and hair to the princess in all her shots. She wears the white nightgown.
15 s, wide 16:9 matching the original
The soundtrack from the trend example
You dance
One photo. Wan 3.0 makes the five meme shots directly from your photo, in the velvet tuxedo.
10 s, vertical 9:16
"Tip Toe Wing In My Jawwdinz"
You and a friend
Two photos, dancer first. Same as You dance, and your friend becomes the person in the white nightgown who gasps.
10 s, vertical 9:16
"Tip Toe Wing In My Jawwdinz"
Both Original clip versions cost more than the others because they recreate 15 seconds of the actual clip, which can take several minutes. The 2-person and Whole body options cost the same as 1-person. Each version shows its exact credit price on its card and the Generate button before you commit.
Which part of the clip is it? The 15 seconds starting when the princess turns and gasps through the straw becoming gold: the peek, the shush, her shock, the tiptoe dance, the shoes in hay and finger raised, ending with her "Oh my God". Our edit skips three very brief close-ups (his face, her eyes, and her face with his hand in her hair) because our first model handled faces poorly in those moments.
Why is Original clip wide? It preserves the original film's frame. The princess sits to one side in wide shots, so a vertical crop would remove her.
How to include the "tip toeing in my jordans" soundtrack
The meme's song is Riff Raff's "Tip Toe Wing In My Jawwdinz" (2014). People reference its hook as "tip toeing in my jordans". Smew AI does not put the Riff Raff song on Original clip. That version keeps the soundtrack from the trend example. You dance and You and a friend use Wan 3.0 and include Tip Toe Wing In My Jawwdinz. The song belongs to its owners.
Original clip does not include the Riff Raff song. To add it when you post:
TikTok: upload your MP4, tap Add sound, search "tip toe wing in my jawwdinz" and select RiFF RAFF's track. Open Volume and reduce the original audio to 0 so only the song plays.
Instagram Reels: upload your footage, tap the music icon, search the same title, then lower your clip's audio to 0 using volume controls.
YouTube Shorts: add the track from the Shorts audio library in the editor, where available in your region.
The dance movements won't sync precisely to the song's rhythm, and the original doesn't either. That loose, off-rhythm feel contributes to the humor.
Common issues and solutions
We ran these prompts through five tests using three AI-generated people before release. All five iconic meme moments appeared correctly ordered in each run: the hush, princess's reaction, tiptoe toward camera, shoes in hay, and broad grin. Here are actual issues we encountered:
What goes wrong
What the prompt does about it
If you still see it
Dancer's face looks less like you in final close-up
The still preserves your face, hair and beard, while the video prompt emphasizes "keep the dancer's face" for each shot
Smaller faces in stills drift most. In one test a beard thinned to stubble in the final close-up. Use sharp, bright photos, or generate again.
Your hair becomes wild and curly
The prompt requests your natural hair, "a little wild and messy", matching the original dancer
This affected a straight-haired woman in testing. Edit the photo prompt and delete that phrase to keep your usual style.
Your glasses vanish
The prompt preserves face, hair and facial hair
One test removed a man's glasses. Edit the photo prompt to add "keep their glasses".
The barn appears in daylight instead of nighttime
The prompt specifies night with warm lighting on hay
One of five runs produced daylight. Generate again; each run differs.
The princess receives your face
The prompt describes her as a separate person with distinct features
Didn't occur in our testing. If it happens, generate again.
Original clip and Original clip, 2 people replicate the actual film using Seedance 2.5 since 13 October 2026. Previously they used MiniMax H3 Max Recast, which frequently lost beards or retained the original man's pale appearance in close-ups. In our Seedance 2.5 tests with AI-generated people, a cat and a hooded man, the photo's face appeared on the dancer in all shots, including the "shh" close-up and every grin, preserving glasses, beards and curly hair, and in the 2-person format the princess received the second photo's face and hair throughout all scenes while keeping the white nightgown. Limitations observed: in one 2-person run the second half shifted by roughly one second, and glasses disappeared from two close-ups. Each run varies, so regenerate if a shot appears incorrect.
Creating the Rumpelstiltskin dance in other tools
Three common approaches exist, each serving different purposes.
Route
How it works
Good for
Green-screen formats
The original dancer appears isolated on green background, which you overlay onto your footage or images in an editor.
Placing the original character in your home, workspace or bedroom
CapCut "tip toeing in my jordans AI trend" templates
Select the template in the CapCut mobile app and upload a photo. Many TikTok guides from 6 October demonstrate this.
Quick results if you already use CapCut for editing
Smew AI, on this page
Upload a single face photo and tap Generate. Original clip inserts you into the actual footage with Seedance 2.5, and Original clip, 2 people includes a friend as princess. You dance and You and a friend use Wan 3.0 to make the hush, the reaction and the tiptoe routine directly from your photo.
Becoming the dancer yourself, single tap, no prompts, no watermark
Original clip replicates the source footage precisely, placing your face on the dancer. The other two formats generate new footage featuring you, so their shots approximate the meme without frame-by-frame accuracy.
FAQ
Rumpelstiltskin Dance trend questions
What is the Rumpelstiltskin AI trend?+
A viral meme from October 2026: AI-generated footage resembling vintage fairy-tale cinema, featuring a peculiar character in glossy velvet who hushes the camera and tiptoes through a hay barn while a princess reacts. Set to "Tip Toe Wing In My Jawwdinz", it spread across TikTok and X. Now users insert their own faces as the dancer.
Is the Rumpelstiltskin 1987 tiptoe footage genuine or AI?+
It's AI-generated. AI creator Tatyana Tereshchenko (@stroinaya) shared it on Instagram 27 August 2026. It's not from the actual 1987 film *Rumpelstiltskin*, a Cannon musical starring Billy Barty and Amy Irving, nor from vintage Russian cinema. The "Rumpelstiltskin (1987)" caption on reposts caused widespread confusion.
Who created the original Rumpelstiltskin AI footage?+
Tatyana Tereshchenko, posting as @stroinaya on Instagram. She released it 27 August 2026 with @neuroferma and @anna_glinskaya, captioned about creating fantasy film without cameras or actors. She hasn't revealed which AI tools were used.
What song plays in the Rumpelstiltskin meme?+
"Tip Toe Wing In My Jawwdinz" by Riff Raff, released 2014. People reference its hook as "tip toeing in my jordans", giving the meme its name. On TikTok the song appears in the audio library under RiFF RAFF.
What does the Rumpelstiltskin meme represent?+
Nothing profound. The eerie character represents silly dancing, sneaking around ("creeping downstairs at 2am for a lil treat") or odd reactions to situations. Swapping a friend's face creates a roast they'll want to share.
How do I create the Rumpelstiltskin dance featuring myself?+
This page functions as a Rumpelstiltskin AI generator. Tap "Create your own", upload one clear face photo and tap Generate. With **Original clip** (default), Smew AI places your face on the dancer in actual footage, or your full image with **Whole body**. With **You dance**, Wan 3.0 makes the tiptoe clip directly from your photo: the hush, the reaction and the routine. No prompt required, and your clip appears in your Library as MP4.
Can I include a friend in the video?+
Yes, two ways. **Original clip, 2 people** keeps actual footage and places your friend's face on the princess: upload dancer first, princess second. **You and a friend** creates new footage: upload dancer first, observer second, and the second person becomes the nightgown-wearing observer who gasps. Only use friend photos with their permission.
Can I feature a celebrity or streamer?+
No. While many viral swaps featured famous people, fake videos of real individuals can violate platform policies and personal rights, and video models frequently reject recognizable faces. Use yourself or consenting friends.
Does my video include the "tip toeing in my jordans" song?+
No. Original clip keeps the source clip's score, which is not the Riff Raff track. You dance and You and a friend use Wan 3.0 and include Tip Toe Wing In My Jawwdinz. The song belongs to its owners.
Is there a Rumpelstiltskin green screen version?+
Fans released green-screen cut-outs of the original dancer, allowing placement in front of refrigerators, offices and TV programs. Smew AI offers something different: it generates new footage with you as the dancer, eliminating the need for green screens or editors.
Can I export it as a GIF?+
Smew AI produces a vertical MP4 with audio, not a GIF. To create a GIF of yourself tiptoeing, extract two or three seconds from the MP4 using any GIF converter.
How long is the video and how long does generation take?+
Original clip and Original clip, 2 people run 15 seconds, wide 16:9 matching the source film, at 720p with its audio, requiring up to several minutes. You dance and You and a friend run 10 seconds, vertical 9:16, with audio, completing in a few minutes. You can navigate away during processing, and completed footage waits in your Library.
Does the Rumpelstiltskin video require credits?+
Yes. Each video costs credits, with exact pricing shown on the Generate button before purchase. All four formats use Seedance 2.5. Both Original clip formats cost most because the model processes 15 seconds of actual footage plus 15 seconds generated. Failed videos refund credits.
Can I share my Rumpelstiltskin video on TikTok or Instagram?+
Yes. The MP4 has no watermark and includes original audio, so you can upload to TikTok, Reels or Shorts as-is or with trending audio added. Keep platform AI labels active, and use #tiptoeinginmyjordans, the most frequent trend tag in TikTok clips we reviewed.
No watermark. A failed video gives its credits back. Secure checkout by Stripe.
Rumpelstiltskin Dance video128 credits · ready in a few minutesCreate your own
Cookie Settings
We use cookies to improve your experience, analyze site usage and support our marketing. You can accept or reject non-essential cookies; essential cookies needed for sign-in always stay on.