Ghost mannequin (sometimes called “invisible mannequin” or “hollow man”) is the highest-volume AI fashion workflow on the planet. It is also one of the most misconfigured. Three settings decide whether your output looks like a clean catalog shot or a melted t-shirt — and most teams misset all three.
What the ghost mannequin shot is for
Ghost mannequin shows the garment with body shape and drape, but no model in frame. The clothing looks like a person is wearing it, even though the person is invisible. Used for catalog listings, Amazon main images, wholesale linesheets, and PDPs where the brand wants to show fit without a face.
It is the right tool when the product is the subject and the person would distract from it — or when you do not want casting, model rights, or a specific face attached to a SKU.
Setting 1: the source photo
The most common failure point. Source photos for ghost mannequin should be either flat-lay (garment laid flat on a clean surface with the cut visible) or on-mannequin (garment on a real mannequin, photographed from a clean angle). Folded garments, stacked product piles, or garments crumpled into a basket produce melted output.
Best results: flat-lay or on-mannequin, even light, no shadows across the garment, no other items in frame. Apiway accepts files up to 50 MB. Resolution above 1500 px on the long edge.
Setting 2: garment type
Ghost mannequin works best for structured garments: shirts, jackets, dresses, blazers, knitwear with body, structured pants. It works less well for unstructured fabrics that need a body to give them shape: very thin slip dresses, drapey scarves, loose camisoles, beachwear cover-ups.
The simple test: if you laid the garment on a flat surface and a friend could tell what it is, ghost mannequin will work. If the garment looks like a puddle of fabric on a flat surface, ghost mannequin will too.
Setting 3: aspect ratio and framing
For Amazon main images: 1:1 square. For Shopify catalog: 4:5 portrait. For wholesale linesheets: 4:5 or 1:1 depending on retailer convention. The aspect ratio is locked at the start of the catalog work and does not change between SKUs.
Framing should fill the frame: the garment occupies at least 85% of the image area. Tight crops work better than loose ones for catalog work because the garment detail is what shoppers came for.
How Apiway handles ghost mannequin
On Apiway, Ghost mannequin is a dedicated template. Upload the source image, pick the aspect ratio, and submit. The post-processing pipeline forces a true #FFFFFF background. (Detail: why we re-composite onto pure white.) Output drops directly into Amazon, Shopify, or wholesale linesheets.
Cost and batch shape
Ghost mannequin runs about one credit per shot. For a 200-SKU catalog, that is roughly $2 in compute plus the operator time to upload the source files. The batch can run overnight and ship in the morning.
(Pricing recap: one credit equals one cent.)
When ghost mannequin is the wrong tool
For lifestyle imagery, ad creative, and any context where a human is part of the brand story, ghost mannequin is too clinical. Use White Studio for on-pure-white on-model shots, or a creator photo set with Virtual try-on for lifestyle.
Run a single garment first
Pick one of your worst current catalog shots. Lay the garment flat, photograph it on your phone, and run a single ghost mannequin pass. Free accounts ship with 100 one-time credits — enough to test on dozens of SKUs.