Put your photo first, add one photo per piece, write which piece goes where, and Lovart ME dresses you in the whole look while your face and pose stay.
Last checked 2026-11-04
Google's cutting-edge image generation model
OpenAI's GPT-4o image generation model
ByteDance image generation and reference-based editing in 2K, 3K, or 4K
Flux Kontext standard model with balanced performance
GPT Image 2.5 Flare
When this option is enabled, the output image may be selected by AI Generate and published to the Explore.
Four looks built here from original photos: five pieces into one outfit, three outfits from a six-piece capsule, a flat lay put on a model, and a date night look from one preset.
The people and clothes below are original generated adults and original pieces with no brands. Each result was made once on November 4, 2026 with GPT Image 2.5 Flare through the project's sample tools, the same image-to-image route the workbench above uses; Use this preset loads the preset and piece map into the workbench, then you add your own photos.
Full Look Builder 
Capsule Week 
Store Lookbook 
Date Night 
Load your photo as image 1, add each piece as its own image, write one line that maps pieces to images, then press Build the look.
Image 1 is always you: one person, standing straight, facing the camera, arms slightly away from the body, even light, plain background. Full body when trousers and shoes are in the look.
Every slot after your photo is one piece: a top, trousers, shoes, a bag, a jacket. The default model takes up to 16 images in total, so you can lay out a whole look; it runs in the browser on desktop or phone, with nothing to install and no cutting out.
In Piece map, write one line such as “sweater from image 2, trousers from image 3, sneakers from image 4”. The keep-list for your face, pose and background is already part of the effect.
Pick a Preset and press Generate; the cost shows on the button first. Compare the result with your photo; if one piece is off, run again with only that piece changed.
An AI outfit changer gets every piece right when you say where each one comes from: "sweater from image 2, trousers from image 3, sneakers from image 4".
“AI outfit changer” usually means the whole look at once: top, bottom, shoes and accessories. “AI clothes changer” is the wider name and often means one piece. Both run here; the difference is how many piece photos you add and how clearly the piece map names them.
Upload the result as image 1 and only the new piece as image 2, then write “replace only the trousers with image 2”. The rest of the outfit stays.
Shoes are the hardest piece to get right: give them their own photo from a three-quarter angle and name them in the map. Say which hand holds a bag, as the sample did with the tote.
Pick Full Look Builder for one outfit, Capsule Week for three outfits from six pieces, Store Lookbook for flat lays on a model, or Date Night for a ready look.
The default: every piece in one outfit. Best with one photo per piece and a full piece map.
Pick a preset →Three outfits from the same pieces, side by side, for planning a week or a trip from one small wardrobe.
Pick a preset →A flat lay, hanger or ghost-mannequin shot on a model. Before you publish, check colour, print placement, sleeve and hem length, neckline and pockets against the real garment.
Pick a preset →A ready evening look from your photo alone, or built from your own piece photos.
Pick a preset →Google's Try it on and Doppl are free in the US for listed items; the AI clothes changer here takes any piece photos, up to 16 images, and builds the whole outfit.
| Route | Where the clothes come from | Pieces at once | What you prepare | Limits |
|---|---|---|---|---|
Google Shopping Try it on (free, US) | Eligible dresses, tops, pants, jackets and shoes in store listings | One listed item | A full-body photo | US only, listed items only; links back to the retailer |
Doppl app (free, US) | Outfit photos or screenshots | A look per photo | A full-body photo, the iOS or Android app | US app from Google Labs (June 26, 2025); short animated previews |
AI clothes changer on Lovart ME (uses credits on this site) | Any piece photo: stores, flat lays, your own wardrobe | Up to 15 pieces after your photo | Your photo, piece photos, a piece map | Credits per run, shown on the button |
Every route shows style and colour; size still comes from the store size chart.
Build the look →Use one person standing straight in even light, full body when trousers and shoes are in the look, and one flat, front-facing photo per piece.
✓ One person, upright, facing the camera · ✓ Arms slightly away from the body, hands visible · ✓ Even light, plain background · ✓ Head to feet for trousers and shoes
✕ Crossed arms, bags over the body, mirror selfies with the phone in front
✓ The whole item, front facing, unfolded · ✓ Straps, sleeves, hem and buttons in view · ✓ Plain background, one item per photo
✕ Folded piles, a piece worn by another person, heavy shadows
Run again with only that piece changed and name it first in the piece map. For a small fix, such as a wrong button colour or a crop, open the result in PhotoCraft, the free editor here.
The AI clothes changer reads your photo and every piece photo, then redraws only the outfit, so face, hair, pose and background stay and you end up fully dressed in the new look.
It exchanges one complete outfit for another; the person stays fully dressed in every result. Models and the credit cost follow the site's current configuration and show on the button before a run. Before you upload, check how photos are stored and shared in the privacy policy, and upload only yourself or adults who agreed.
Read the privacy policy →Lay out a lookbook page in PhotoCraft, animate the outfit in image to video, or make product shots for the store.
Put the Capsule Week looks on one page, add names and prices in PhotoCraft, export for the shop or a post. PhotoCraft is free.
Send the finished photo to Animate the look on Image to Video for a short turn or walk.
Need the piece on its own? Make clean shop images with Product shots.
Change your haircut too → AI Hairstyle Changer. Try a design on your skin → AI Tattoo Generator.
It swaps the outfit in your photo for a new one built from your piece photos or a description, and keeps your face, hair, pose and background. Here your photo goes first and every piece follows as its own image.
One person standing straight to the camera, arms slightly away from the body, hands visible, even light and a plain background. Use a full-body photo when trousers and shoes are part of the look.
Each item whole, facing front, unfolded on a plain background, with straps, sleeves, hem and buttons in view, one photo per item. A store product photo works.
Yes. Save the product photo and add it as image 2 after your photo. The result shows style and colour; for size, use the store’s size chart.
Write the piece map first (“sweater from image 2, trousers from image 3”), and the effect adds the keep-list: face, hair, expression, body shape, pose, hands, light, background and camera angle.
Read the privacy policy for how uploads are stored and shared; transfers are encrypted and you can delete your account at any time. Upload only yourself or adults who agreed. Privacy policy.
Only the clothes. The person keeps face, hair, body and pose, and is fully dressed in the new look in every result.
One person per run works best. Crop each person into their own photo first, run them one at a time, then put the results side by side in PhotoCraft.
Tops, dresses, jackets and trousers come out best. Shoes are the hardest piece: give them their own clear photo and name them in the piece map.
Yes. Upload a photo of the piece in the colour you want, or write it in the piece map, for example “the same sweater in navy”.
Yes. Any piece photo works, wherever it comes from; put each one in its own slot and say which image holds which piece.
Pick Capsule Week, add your photo and the pieces (one flat lay or one photo per piece), and you get three outfits from the same pieces side by side.
Close enough to judge style, colour and how pieces go together; it does not measure your body, so fit and size still come from the store size chart.
Sources: Google Shopping Try it on, TechCrunch on Doppl, FTC consumer advice (checked 2026-11-04).