Flat Lay to Model With AI: How to Shoot the Flat Lay So Length and Fit Survive

Turning a flat lay into an on-model photo with AI works for some garments and fails for others. How to shoot it, what to add, and what to check.

A flat lay is the cheapest product photo there is: a table, a phone, a window. Many sellers already have hundreds of them. An AI flat lay to model tool takes that photo and draws the garment on a generated model. It works well for some garments and badly for others, and the difference is predictable. This guide covers which garments convert cleanly, how to shoot the flat lay for the AI, and what to check in the result.

What the AI has to guess from a flat lay

A flat lay shows the garment from above with no body inside it. The AI has to infer three things the photo does not show directly:

• Length on the body. A dress laid flat has no knee or ankle next to it, so the hem can land anywhere. This is the most common error.
• Drape and volume. A wide-leg trouser and a straight trouser can look almost the same when flat.
• The back. A flat lay shows one side. Whatever is on the other side is invented unless you show it.

T-shirts, sweatshirts, knits and shirts convert well because their shape on the body is close to their shape on the table. Long dresses, wide trousers, pleated skirts and anything with a structured fit convert worse. For those, a hanger or shop mannequin photo gives better length and fit.

How to shoot a flat lay for AI

1. Shoot from directly above, phone parallel to the table. An angled shot makes the hem look shorter than the neckline.
2. Use a plain surface that contrasts with the garment: white sheet for dark fabric, grey or wood for white fabric.
3. Daylight from a window, no flash. Hard shadows get read as folds, seams or darker fabric.
4. Lay the garment the way it will be worn: buttons closed, zipper up, sleeves straight down the sides, not folded across the chest.
5. Smooth it, but do not stretch it. Pulling the hem sideways makes the garment read as wider.
6. Fill the frame. If the garment takes a third of the photo, prints and small closures get lost.
7. Take a second flat lay of the back if it has any detail: a zipper, a print, a cut-out, a bow.

Tell it the length

Because a flat lay carries no body reference, length is where you help the most. If the tool lets you choose the length or describe the garment, say it plainly: cropped, hip length, knee length, midi, ankle length. A measuring tape laid next to the garment does not help; the AI does not read measurements from a photo reliably.

What to check in the result

• Hem length: compare with the real garment on a person or a mannequin.
• Closures: the number and position of buttons, and whether a zipper is open or closed.
• Print placement: a chest print should stay on the chest, not drift to the stomach.
• Sleeve length: flat sleeves are often drawn longer than they are.
• The back view: if you did not upload a back photo, treat any back frame as invented.

If a detail is wrong, generate again before editing. Results vary between runs, and a second run often fixes a single detail. If the same error repeats, the cause is usually in the photo: a shadow at the hem, a fold at the waist, or a sleeve tucked under.

Flat lay or hanger?

If you are shooting new stock anyway, a hanger against a plain wall is usually the better input: gravity shows the real length and drape. Keep flat lays for knits and tops, and for the stock you already photographed. The same principles apply to hanger and mannequin photos; see our guide on how to photograph clothes with your phone.

vCabin accepts flat lay, hanger and shop mannequin photos, up to three views per garment. A Normal frame costs 10 credits at 2K, a Pro frame 35 credits at 4K, and each takes about 60 seconds.