Skip to content
Ghost Mannequin Tool

Remove a Person From a Model Photo and Keep the Garment as a Flat Lay

If you only have a photo of a model wearing the garment, upload it here: Flat Lay and Worn are already selected. Generate, then check the parts the person covered, such as under the arms or behind hair. Output is 2048 px on the short edge in 1:1, 4:5 or 3:4.

Pick your photo below, then sign in with email or Google to create a watermarked Standard preview. A new account on this site includes 2 trial credits; each new image uses 2. Unlock the existing preview for $0.50 without buying more generations.

Standard · 2 credits per generation

2K source for finer fabric and print detail. Check each result against your garment.

Download: 2048 px on the short edge.

  1. Photo
  2. Settings
  3. Result

Works best

  • One garment, fully in frame
  • Straight-on, evenly lit
  • Plain background

Avoid

  • Cropped hems or sleeves
  • Busy backgrounds or several garments
What does your photo show?

We only keep the garment. The person in your photo won't appear in the result. For worn photos, a front view with the whole garment visible works best.

Extra photos (optional)

More photos of the same garment help the AI get hidden details right. They don't cost extra.

Inside of the collarShows the real neck label, or that there is none
Another angleBack or side, for details the main photo hides
Background
Aspect ratio
Padding

Next generation settings

Framing reference only. These settings apply to the next generation and do not edit the current result.

Output: 2048 × 2048 px · JPG

Sign in with email or Google. New accounts can try 1 Standard image (2 trial credits). Watermarked preview; unlock this already-generated image for $0.50 via PayPal. Trial credits are granted after email verification; Google sign-in counts as verified.

Uses 2 credits per image. A failed image gives the credit back.

  • Private storage: your photos are not publicly displayed
  • Failed images give the credit back
  • 2048 px downloads in 1:1, 4:5 or 3:4
  • Packs from $2.99 · no subscription
Ecru embroidered linen tunic, Worn photo beforeWorn
Flat lay result of the same Ecru embroidered linen tunicFlat lay
D20FEcru embroidered linen tunicStandard mode · 2K · Flat lay · Light wood · 4:5The person is removed, but the ivory is warmer, linen slubs coarser and the embroidered sprig larger with redrawn leaf angles and stitches. All five leaves and four front buttons remain. Sleeves and hem are rearranged and side slits lengthened; the hidden inner collar is inferred by AI.Attempt 1 of up to 3
Ecru embroidered linen tunic, Worn photo beforeWorn
Flat lay result of the same Ecru embroidered linen tunicFlat lay
M12Ecru embroidered linen tunicHistorical Basic example · 1K; new generations use Standard 2K only · Standard has stronger linen texture contrast; basic is softer. Embroidery and side-slit shapes differ, but detail improvement is not obvious at web size. Both retain five leaves and four buttons.The ivory is warmer and the linen weave redrawn. The sprig, leaf angles and stitches are reshaped, while five leaves and four buttons remain. Sleeves and hem are rearranged and side slits lengthened. The hidden inner neckline is inferred from the worn photo.Attempt 1 of up to 3

Compare the original photo and result, including the neckline, print and fabric.

Have a similar clothing photo?

Try your own garment. One clear main photo is enough to start; extra reference photos are optional.

When This Is the Photo You Have

Resellers photograph a piece while trying it on. Small brands get photos back from a friend who modeled the collection. Wholesale suppliers send on-model shots only. In all of these cases the garment is there, but a marketplace or a clean collection page needs it on its own.

Cutting a person out by hand leaves arm holes, a neck and hair in the collar. Here the AI removes the person entirely and redraws the garment laid flat and seen from above, so it looks like a proper flat lay rather than a cut-out.

Which Worn Photos Work

Works well

A front view, standing straight, arms relaxed at the sides, with the whole garment visible from collar to hem.

Harder

Crossed arms, hands in pockets, a bag strap or long hair over the garment. The AI has to guess what's underneath, and that part may not match.

Avoid

Side or back poses, sitting, several people in one photo, or garments that are cropped at the edge of the frame.

How It Works

  1. 01

    Upload the worn photo

    Choose Worn under "What does your photo show?". JPG, PNG or WEBP, at least 800 px on the short side.

  2. 02

    Pick the look

    Choose a background, aspect ratio, padding and whether you want a soft shadow under the garment.

  3. 03

    Check and download

    Inspect the preview for any remaining face, body or hair, then check the garment details. After an image unlock or credit-pack purchase, download the same result as a JPG with a 2048 px short edge.

What to Check Before You List It

Hidden parts

Anything the person's arms, hair or a bag covered is drawn by the AI: often the sides of the body, the inside of the sleeves and the back of the collar. Compare those areas with the garment.

Size and fit

A flat lay shows the garment's shape, not how it fits on a body. Add measurements to the listing, such as pit to pit and length, taken from the real garment.

Tags and labels

The result can show a neck tag that wasn't in your photo. Remove anything your garment doesn't have before you publish it.

Other people's photos

Only use photos you took or have permission to use. Removing the person doesn't change who owns the photo.

Privacy and Honesty

Only upload photos of people who agreed to it, and never photos of children. Your uploads are private and deleted automatically after 30 days.

Because the garment is redrawn laid flat, parts hidden by the pose are AI-generated. Compare the result with the real piece before you list it, especially the sleeves and the back.

When your supplier only sent model photos

If your supplier only sent model photos, you can turn an on-model clothing photo into a flat lay here. Use a photo you have permission to edit, choose Worn as the input and Flat lay as the output. A front view with the full garment visible works best. The AI redraws areas hidden by arms, hair or the pose, so compare the neckline, sleeves, hem and print with the actual product before using the image in a listing.

Remove Person FAQ

Will the person's face or body appear anywhere?

The intended output is the garment alone, laid flat. AI can leave artifacts, so inspect the preview for any face, body or hair before downloading or publishing. Your original upload is stored privately.

Can I get a ghost mannequin from a worn photo instead?

Yes. Open the ghost mannequin generator and choose Worn for a dimensional view without the person. Use the flat lay tool for a top-down view instead. Hidden areas are generated by AI, so compare the result with the actual garment; it is not a measurement of fit.

Does it work with a mirror selfie?

Often, yes, if the phone doesn't cover the garment and the whole piece is in frame. A photo taken by someone else, straight on, works better.

Can it keep the model and change the background?

No. This tool always removes the person and makes a flat lay.

What does the trial include, and what do I pay to download?

Sign in with email or Google. New accounts can try 1 Standard image with 2 trial credits and inspect a watermarked preview. Unlock that existing image for $0.50, or buy a one-time credit pack for more generations and an unlocked trial image. The $0.50 unlock adds no generation credits; neither option starts a subscription.

How many credits does converting a worn photo use?

Standard uses 2 credits per generated image. Each new version costs another 2 credits. Failed generations return the charge automatically. Extra photos of the same garment can be added in Single item processing without an extra credit charge.