Walk Outfit Change AI Video Prompt: This Is My Ex
Walk outfit change AI video prompt for Seedance 2.5. You see your ex, say one line and your outfit changes as you walk away. 15 seconds, two looks.
Have it running in three steps
- 1
- 2Open Vosu
- 3Paste it and generate
What you'll need
30 min- Vosu
- Claude
This walk outfit change AI video prompt is a small story in 15 seconds. You walk along a park path in a comfy flannel shirt and sweatpants. You stop, look to the side and say quietly: “Oh, this is my ex.”
Then you walk on, and your clothes change while you walk. Tiny fabric petals flip over from your shoulders down to your boots. A few steps later you wear a knit sweater, a plaid mini skirt and over-the-knee boots, and your hair is in a braid. The ex never shows up on screen. The viewer only sees your face.
You need one photo of yourself and three prompts. Two make the pictures of your first and second look. The third makes the video.
The whole process
- Pick one clear photo of your face, in good light.
- Make your first look with the first prompt. This is @Image1.
- Make your second look from the first one with the second prompt. This is @Image2.
- If your clothes are different from ours, adapt the video prompt in Claude.
- Open Vosu, pick Seedance 2.5, add both looks, set 15 seconds and 9:16, paste the prompt.
What happens in 15 seconds
| Time | What happens |
|---|---|
| 0–3s | She walks toward the camera in the flannel shirt and grey sweatpants |
| 3–5s | She stops, looks to her right, her eyebrows go up. “Oh, this is my ex.” |
| 5–6s | A quick smirk, she turns forward and walks on |
| 6–9s | Petals flip from her shoulders down. Sweater, skirt, boots, bag and braid appear |
| 9–15s | She walks on with a longer stride, touches the braid and smiles |
Make your first look
The first look is the outfit you start in. It is casual on purpose, so the change has somewhere to go.
Open an image model that takes a reference picture, for example on Vosu. Add your photo, set the shape to 4:5 and paste this prompt. Change the clothes in the CLOTHES block to anything you like.
Use Image 1 as the only source for the person’s identity. Create a fashion lookbook photo of this person in a casual autumn outfit.
IDENTITY — IMAGE 1: Keep exactly: face shape and bone structure, eye shape and eye color, eyebrows, nose, lips, skin tone, freckles and moles, age, body build, hair color and hair length. Do not beautify, slim or redesign the person. The same person must be recognizable in both parts of the image.
HAIR: Worn straight and loose, with a centre part.
CLOTHES:
- an open forest-green plaid flannel shirt with rolled sleeves
- a plain white crew-neck t-shirt
- light grey jogger sweatpants with a drawstring
- olive green high-top canvas sneakers
- one thin gold necklace with a small coin pendant
LAYOUT: Portrait 4:5 canvas, plain off-white studio background. Left side: the person full-body, standing, facing the camera, hands in pockets, head to shoes fully visible. Right side: a large framed close-up of the face and shoulders in the same outfit, with a thin border.
PHOTOGRAPHY: Real studio photograph, 85mm lens, soft even light, true-to-life color, real skin texture with pores, real fabric texture.
CONSTRAINTS: One person only. No text, no labels, no numbers, no watermark, no illustrated or 3D look.
This is the first look we used.
Make your second look
Now make the outfit you change into. Add the first look as the reference, not your photo. Then the face, the light and the layout stay the same, and only the clothes and the hair change.
Use Image 1 as the reference for the person’s identity, the layout, the background and the lighting. Keep the same person, the same framing and the same off-white studio background. Change only the outfit and the hairstyle.
IDENTITY: Keep exactly: face, eye color, freckles, skin tone, age, body build and hair color from Image 1. Do not beautify or redesign the face.
HAIR: The same hair, now in one loose side braid over the left shoulder, with soft strands around the face.
NEW OUTFIT:
- a forest-green chunky ribbed knit crew-neck sweater
- a grey, cream and dusty-rose plaid flared mini skirt
- taupe suede over-the-knee boots with a low block heel
- a small blush-pink quilted crossbody bag on a thin gold chain strap, worn across the body
- two thin layered gold necklaces with small coin pendants No logos on the bag.
LAYOUT: Same as Image 1. Left side: the person full-body, standing, facing the camera, head to boots fully visible. Right side: a large framed close-up of the face and shoulders, showing the braid, the sweater and the necklaces.
PHOTOGRAPHY: Same studio photograph style as Image 1: 85mm lens, soft even light, true-to-life color, real skin and knit texture.
CONSTRAINTS: One person only. No text, no labels, no numbers, no watermark, no illustrated or 3D look.
This is our second look. Put the two pictures side by side and check the face. If it changed, make the second look again before you move on.
Adapt the video prompt in Claude
Skip this step if you used our clothes. If you changed any piece of either look, the video prompt will not match your pictures. Seedance then mixes the words and the pictures, and you get the wrong colors or a piece that never appears.
Open Claude. Add both of your look images, paste this message and paste the video prompt from Step 4 under it. Claude reads the clothes from your pictures and rewrites only the parts about clothes, hair and shoes.
I am attaching two images of the same woman. Image 1 is her starting outfit. Image 2 is her final outfit.
Below is a video prompt for Seedance 2.5. It was written for different clothes. Rewrite it for my two images.
Rules:
- Change only what is about clothes, shoes, bag, jewelry and hairstyle. Keep everything else word for word: the timing, the camera, the setting, the dialogue, the petal effect, the constraints and the audio.
- In [References], describe what each image defines, using what you see in my images.
- In [Stage 3], list each piece of the starting outfit and what it turns into, with its real color and fabric, the same way the original does. Keep the line about continuous coverage if the change goes from trousers to a skirt or a dress.
- If the hairstyle does not change, remove the braid part.
- In [Audio], match the footstep sounds to the shoes before and after the change.
- Return only the finished prompt, in English, in one block, ready to paste.
Video prompt:
Read what Claude returns before you use it. Every piece in your second picture should be named in Stage 3.
Set it up and generate
Open Vosu, go to video and pick Seedance 2.5.
Add the first look as @Image1 and the second look as @Image2, in that order. Set the length to 15 seconds and the shape to 9:16. Paste the prompt below, or the one Claude adapted for you. This is the exact prompt that made the video above.
[Generation Goal] One continuous 15-second take, no cuts, no music. A woman walks toward the camera beside an autumn park, stops, spots her ex off-screen to her side, says “Oh, this is my ex.” and walks on while her outfit transforms mid-stride through a quick ripple of tiny feather-shaped fabric petals. She continues walking, pleased with herself. Only the woman appears on screen. [References] @Image1 defines her facial identity, freckles, green eyes, hair colour, straight loose hair and starting outfit. Do not reproduce the studio background, portrait inset, borders, labels or hands-in-pockets pose. @Image2 defines her final outfit, accessories and loose side braid. Do not reproduce its studio background, inset, borders or standing pose. Both references show the same woman. Preserve her face and body proportions throughout. [Setting and Look] Dark overcast autumn day. Wet asphalt path beside a park, fallen leaves, bare branches, muted foliage. Soft grey daylight, subdued green-grey-amber palette, realistic skin and fabric textures, subtle film grain, softly blurred background, light breeze in her hair. The path is empty apart from her. Her ex stays entirely off-screen: no man, passerby, silhouette, reflection or second person. [Camera] The camera faces her at chest height and tracks backward at her walking speed. When she stops, the camera stops with her; when she walks on, it resumes tracking. Medium-full framing from head to just below the knees. Face always sharp. No cuts, zooms, orbits or reframing to show what she sees. [Stage 1 — 0–3 s] She walks toward the camera in the @Image1 outfit, relaxed, arms swinging naturally. End state: still walking, calm expression. [Stage 2 — 3–6 s] Primary event: she stops and turns her head to her right, looking off to the side, away from the camera. Her eyebrows jump up, her eyes widen and her lips part. Still looking to the side, not at the camera, she says the line. Then a quick knowing smirk, she turns her head forward and starts walking again. Dialogue language: American English. She says quietly, surprised, under her breath but clearly audible, with accurate lip sync: {Oh, this is my ex.} End state: walking forward again, smirking, the transformation starting the moment she takes her first step. [Stage 3 — 6–9 s] Continue from the previous stage: she keeps walking, never stops. Primary event: a quick transformation ripple. Tiny feather-shaped fabric petals, each only the size of a grain of rice, far smaller than a fingertip, cover the surface of her clothing in a dense fine texture like fine plumage. Each petal lifts its tip, flicks over and lies flat in the opposite direction, revealing the new material on its underside. The ripple starts at her shoulders and sweeps down her sleeves, torso, waist, hips and legs in a fast continuous cascade, like a hand brushing velvet against the grain. Ahead of the ripple the original clothes stay intact; directly behind it the petals merge into seamless finished fabric with no petal shapes left. Each petal carries the real colours of the clothes: forest-green ribbed knit becomes the sweater, grey-cream-dusty-rose plaid becomes the flared mini skirt, taupe suede becomes the over-the-knee boots. The trousers reshape directly into the skirt and boots with continuous coverage, no exposed underwear. At her hip, tiny blush-pink petals fold into the quilted bag while the chain strap forms along its diagonal. Gold necklaces resolve on the sweater. At the same time her loose copper-red hair sweeps to one shoulder and weaves into the loose side braid. Her face and bare skin stay unchanged, with no petals on her head or face. End state: by 9 seconds her outfit, bag, necklaces and braid match @Image2. [Stage 4 — 9–15 s] Continue from the previous stage: same outfit, same framing. Primary event: she lifts her chin, pulls her shoulders back and walks with a longer, confident stride and a slight hip sway. She lightly brushes the braid with one hand and gives a satisfied closed-lip smile, her eyes slightly narrowed with pleasure. End state: she is walking toward the camera, and the complete look is clearly readable. No further transformation or speech. [Transformation Constraints] Petals are tiny, soft, matte, flexible and fabric-coloured, and hug the body. No large feathers, real bird feathers or fluffy plumage. No metallic scales, sequins, reflections, hexagons, rigid tiles, armour plates, spikes, floating fragments or rings around the body. No smoke, sparks, glow, particle dissolve, crossfade or instant swap. [Continuity] One woman only. Same face, freckles, green eyes, hair colour and body proportions throughout. One uninterrupted shot. No background people. No text, subtitles, labels or logos. [Audio] No music at any point: no score, no soundtrack, no background melody, from the first second to the last. <soft sneaker footsteps on damp asphalt>, <light wind>, <rustling leaves>. Her single line, clearly audible. <a quick soft fabric rustle synced to the petals flicking>. <low-heeled boot clicks on asphalt after the change>. No other voices. No music.
The [References] line also names our face details: freckles and green eyes. Change them to yours, or ask Claude to do it in Step 3.
Why this prompt works
The ex stays off screen. The prompt says it three times: only the woman appears, no man or silhouette, and no camera move to show what she sees. The viewer fills in the ex on their own, and her face tells the whole story.
The camera stops when she stops. It tracks backward at her walking speed and pauses with her for the line. So the framing stays the same, and the change happens on a body the viewer is already watching.
The transformation has a direction. It starts at the shoulders and runs down to the boots. Ahead of it the old clothes stay whole, and behind it the new fabric is finished. The viewer can follow the change with their eyes.
The petals carry the real colors of the new clothes. Green for the sweater, plaid for the skirt, suede for the boots. So the change reads as one outfit turning into another, and the result matches the second picture.
Stage 4 gives the look time to be seen. Six seconds of a confident walk with no more changes, so the viewer sees the whole outfit before the reel ends.
Where it goes wrong
Our first try had two of these problems. The ex walked into the frame behind her, and the petals came out as shiny metal scales. The lines below fixed both.
| What you got | The fix |
|---|---|
| A man walks into the frame | Keep “no man, passerby, silhouette, reflection or second person” and “No background people” |
| The petals look like metal scales or sequins | Keep the [Transformation Constraints] block as it is |
| The petals look long and spiky | Keep “each only the size of a grain of rice, far smaller than a fingertip” |
| The whole outfit swaps in one frame | Keep “No smoke, sparks, glow, particle dissolve, crossfade or instant swap” |
| She says the line to the camera | Keep “Still looking to the side, not at the camera, she says the line” |
| The studio background or the face inset shows up | Keep the “Do not reproduce” lines in [References] |
| Some pieces of the new look are missing | Adapt the prompt in Claude, so every piece in @Image2 is named in Stage 3 |
| Music plays under the scene | Keep both “No music” lines in [Audio] |
Questions people ask
Which AI model makes this video?
We made it with Seedance 2.5 on Vosu. The prompt is written for it: 15 seconds, one take, two reference images and one line of speech.
Can I put myself in it?
Yes. Make both looks from your own photo in Steps 1 and 2. The video takes your face from them.
Can I use my own clothes?
Yes. Change the clothes in the image prompts, then adapt the video prompt in Claude as shown in Step 3. The transformation part must name your real clothes and colors, or the change will not match your pictures.
Can I change the line?
Yes. Change the words inside the curly brackets in Stage 2. Keep it short. One line of under two seconds fits the stop in the timing.
Is it vertical?
Yes, 9:16, ready for Reels, TikTok and Shorts.
Is it free?
The prompts are free. Vosu is a paid service with a trial, so the images and the video use credits from your account.
Ready? Open Vosu, make your two looks and paste the video prompt.
Why this workflow runs here
Made for short-form
Trial available
One place for the video step
Make yours in the next 30 min
You have the whole prompt and every setting. The only thing left is an account, and that takes a minute.
- Free to start
- Prompts ready to paste
- Upgrade only if you want more
Referral link — if you subscribe I earn a commission, at no extra cost to you. More on that.
Related prompts
Effect Free 24 GTR AI Prompt: Swap the Car for Your Dragon
A 24 GTR AI prompt for the AIReel effect. You flash the UNO cards, the scene cuts to a stone mountain, and a dragon stands behind you instead of a car.
Effect Free Fan Blade Outfit Change: The Top-Down AI Prompt
A free Seedance 2.5 prompt. She lies on a rug, a slow ceiling fan blade sweeps over her, and her outfit is new. Five looks, 20 seconds, five reference images.
Effect Free Fit Check Hit by a Car: The AI Video Prompt
A free Seedance 2.5 prompt. She films a street fit check on her phone, shows the outfit from every side, and a car takes her out of frame. 12 seconds, one reference image.