One flat lay in. On-model frames out.
A flat lay is the cheapest useful photo of a garment and the least persuasive one. It shows the thing, not the wearing of the thing. Getting from one to the other has meant a model, a studio and a booking, which is why most catalogs have thousands of flat lays and a handful of on-model shots.
One photograph per garment in
1792 x 2400 delivered files
Up to 60 items by 8 models per job
5 to 7 days to delivery
What a flat lay is good at, and where it stops
Flat lay photography is the garment laid out and shot from directly above, usually on a plain surface, styled to show cut and detail. It is fast, repeatable, cheap, needs no model, and for detail and construction it is genuinely the clearer image.
It stops at fit. A flat lay cannot show drape, length on a body, how a hem falls, or how a jacket sits at the shoulder. In a Stylitics shopper survey, 76% named on-model photography the most helpful format for deciding on a garment, whether the image was human or AI made. That is their research and their finding, not a performance claim from us. Most apparel catalogs cannot supply that format at variant depth, because it has meant a shoot per colorway.
Survey figure from Stylitics shopper research, quoted as market context.
How to shoot a flat lay worth converting
Practical, because a chunk of this traffic is here to learn:
- One diffuse light source, from one side. Overhead flat light kills texture.
- Steam it. Creases read as damage at listing size, and they are the single most common reason a flat lay looks amateur.
- Shoot square to the garment. A tilted camera adds keystone distortion that makes a symmetrical piece look cut wrong.
- Fill the frame with the product. Shopping channels require it, and it also survives being small in a feed.
- Keep the surface boring. Textured backdrops compete with the fabric.
A good flat lay is also a better input. It works from whatever you have, but a steamed, evenly lit, square-on photograph gives a cleaner result than a wrinkled phone snap on a duvet.
The step after generation is the product
One photograph per garment goes in. It does not have to be a flat lay: a hanger shot, a form shot, or the garment worn by someone all work. What comes back is the garment on a model of your choosing, full length, on a consistent set, at 1792 x 2400.
Generation is the first step, not the product. What happens after it:
- The garment is checked. Is the thing in the frame the thing you sent, and is it still the same category? An unguarded generator returns a top as a dress often enough that this check pays for itself.
- Color is measured against your original photograph, as ΔE00.
- Color is corrected on the delivered pixels. On the documented file, 9.03 as returned by the engine, 1.29 as delivered. Across the 240 delivered frames of the 2026-08-18 run, the 220 that carried a measurement had medians of 2.87 returned and 0.68 delivered.
- The receipt is attached. Which engine, which source photograph, how far the color moved.
Six named standards gate the file. One that misses a standard is named, not shipped, and not charged.
Consistency across a catalog, not one good frame
The model, pose, framing and light come from a picked board rather than a sentence in a prompt box, which is what lets item 1 and item 200 read as the same shoot. Nine shot presets cover the standard PDP setups. Up to 60 items by 8 model selections in a single job, 100+ creatives a month, 5 to 7 day turnaround with first output usually on day five.
Is this the same as AI try-on?
Related, aimed elsewhere. Virtual try-on is a shopper-facing tool: someone on your product page seeing the garment on themselves. This is brand-facing production: you getting on-model imagery for the catalog, the feed and the ads. Same underlying capability, opposite end of the funnel. If you searched for AI try on clothes as a brand, this is the page you wanted.
Production is one link
Converting a flat lay is one step. Deciding which garments deserve which treatment, reading what is working, and running the campaign are separate jobs. We cover the chain: creative intelligence, creative strategy, production, influencer marketing at scale and campaign management. A generator covers one link and hands you an image.
How do you turn a flat lay into an on-model photo?
Send one photograph of the garment. The garment is isolated, placed on a selected model, and composed on a consistent set. The output is then checked for garment fidelity, measured for color difference against your original, corrected, and delivered with the measurement attached.
What is flat lay photography?
Flat lay photography is shooting a garment laid flat and photographed from directly above on a plain surface. It is fast and cheap and shows cut and detail clearly, but it cannot show fit, drape or length on a body.
Common questions
- Does my input have to be a flat lay?
- No. Flat lay, hanger, mannequin or worn all work. One photograph per garment is the input.
- Can I choose the model?
- Yes, from a preset model library, and the same selection holds across the whole run so the frames read as one shoot.
- Is the garment changed?
- No. Garment fidelity is one of the six standards, and color is measured against your own photograph with the number delivered alongside the file.
- What resolution do I get?
- 1792 x 2400 on delivered files.
- Is this virtual try-on?
- No. Try-on is shopper-facing on your product page. This is production imagery for your catalog, feed and ads.
- How many garments per run?
- Up to 60 items by 8 model selections in a single job, 100+ creatives a month on plan.
Send five flat lays.
Including the wrinkled ones. You get on-model frames for each, the measured color difference against your own originals, and what it would cost across your real assortment.