5 Product Photography Inputs for AI On-Model Imagery
On-Model accepts flat-lays, outfits, ghost mannequin, on-model, and hanger shots. See all five input types produce the same PDP-quality fashion e-commerce output.
By On-Model Team
Fashion e-commerce brands photograph products in many ways. Flat-lays on white backgrounds. Styled outfit compositions. Ghost mannequin shots. Cropped on-model photography. Garments hanging on hangers. Every warehouse, studio, and supplier has their own format — and until now, most AI product photography tools only accepted one.
On-Model's flat-to-model feature accepts all five. Upload whatever you have, and get the same PDP-quality AI on-model output regardless of how the original was shot. This post proves it: five different input formats, five different garments, one identity, same consistent results.
Our model for this guide is Einar:
Every section below uses the same PDP preset (front-facing, 45° turn, and back view) to show that the output quality and consistency remain identical no matter what you upload. You can browse all available identities in the model library.
1. Flat-lay product photography: individual garments
The most common format in fashion e-commerce. Individual garments photographed flat on a white surface — one item per image, uploaded together as an outfit.
When brands use it: High-volume catalogs, individual product shots, brands with in-house flat-lay photography workflows.
Three separate flat-lay images in, three PDP-ready angles out. The AI composes the individual garments into a complete, coordinated outfit on Einar — with consistent pose, lighting, and background across all three views.
2. Outfit flat-lay photography
A complete styled outfit photographed as a single composition — all garments arranged together on a surface. One image captures the full look.
When brands use it: Editorial lookbooks, social media content, "shop the look" features, styling guides.
One flat-lay composition in, three on-model angles out. The AI reads the full styling intent from the single image — jacket layered over tee, jeans, sneakers — and translates it onto Einar as a complete look.
3. Ghost mannequin photography
The industry-standard "invisible mannequin" or "hollow man" technique. Garments are photographed on a mannequin form that gets digitally removed, leaving a 3D garment shape floating in space.
When brands use it: Brands with established mannequin photography workflows, product detail pages that need to show garment structure and shape.
Ghost mannequin inputs actually give the AI extra information about garment structure — how the fabric drapes, where seams fall, how the collar sits. This 3D context helps produce especially accurate results.
4. On-model photography (cropped)
Garments already photographed on a real person, but with the model's face cropped out or not visible. This is common when brands have existing model shots but need a different identity, or when suppliers provide anonymous model photography.
When brands use it: Updating existing on-model imagery with a new identity, processing supplier-provided model shots, brands transitioning from traditional to AI product photography.
The system preserves the garment's real-world fit and drape from the body in the input photo, then places it on Einar with the requested PDP poses. The original model's body informs the result but Einar's identity takes over completely.
5. Hanger shot photography
The simplest photography format: a garment hanging on a hanger, showing natural gravity drape. Minimal production — just hang it up and shoot.
When brands use it: Fast-fashion catalogs, wholesale product listings, showroom photography, quick-turnaround product launches where flat-lay styling isn't practical.
Even this minimal-effort input produces full PDP-quality output. The AI extracts the garment's color, pattern, material, and silhouette from the hanger shot and generates three complete on-model views — no flat-lay styling required.
Already have a mix of photography formats across your catalog? Upload them all. On-Model handles the differences so you don't have to reshoot anything.
Which input format should you use for AI product photography?
All five formats produce the same quality output. The best choice depends on what you already have:
- Starting fresh with no product photos? Flat-lay individual garments are the fastest to shoot and give the most control over each item.
- Have a styled lookbook? Upload outfit flat-lays directly — the AI reads the full styling intent.
- Already using ghost mannequin photography? Keep your existing workflow — ghost inputs give excellent results with bonus 3D structure information.
- Have on-model shots that need a new identity? Upload them as-is — the garment fit transfers to the new model.
- Need speed over styling? Hanger shots are the fastest input to produce — hang, shoot, upload, done.
The key factor isn't which format you use — it's input quality. Well-lit, high-resolution images in any format will produce the best results.
What's next
Ready to try your own product photos — in any format?
Start generating — sign up, upload whatever product photography you have, and see consistent on-model results in minutes.
Explore more guides:
- Flat-to-Model guide — step-by-step tutorial for the flat-to-model workflow
- Input formats for e-commerce — deep dive into flat-lay and ghost mannequin inputs
- Scaling product photography — batch workflows for processing thousands of SKUs
- Flat-to-Model presets — using preset categories for consistent multi-style output
Read Next
Flat-Lay to On-Model: How AI Turns Product Photos Into Fashion Imagery
Still using flat-lay or ghost mannequin photography for your fashion catalog? Learn how AI converts flat-lays, individual garments, and ghost mannequin photos into professional on-model imagery — no photoshoot required.
Flat-to-Model Presets: One Outfit, Every Style
Use On-Model's preset categories to generate PDP, lifestyle, editorial, and social imagery from the same flat-lay — plus custom presets for your brand.
Scaling Product Photography: From 10 SKUs to 10,000
Process your entire product catalog with AI on-model imagery. Batch workflows, consistent identity, and the economics of scaling from dozens to thousands of SKUs.