Forty suppliers. Forty different backgrounds. One grid.
A product grid looks cheap when every tile has a different background, a different crop and a different exposure. Fixing that by hand is a week of somebody’s life per season. Fixing it as a pipeline is four API calls per SKU.
as supplied
after the pass
Four calls per SKU, in this order
The order matters more than the tools. Cutting out before upscaling means you are not paying to reconstruct a background you are about to delete.
- 18 cr
remove_backgroundIsolate the product. Soft alpha, so shadows and glass edges survive.
- 2~10 cr
edit_imageOptional: place it on your standard ground with a consistent shadow.
- 315 cr
upscale_imageClear the marketplace resolution minimum and make zoom usable.
- 41 cr
compress_imageWebP for the grid so a 60-tile page does not weigh 40 MB.
const batch = skus.slice(0, 20).map((sku) => ({
tool: "remove_background",
args: { image: sku.supplierPhotoUrl },
}));
const res = await fetch(
"https://api.imagemcpserver.com/playground/multicall",
{
method: "POST",
headers: {
Authorization: `Bearer ${process.env.IMAGEMCP_API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({ requests: batch }),
}
);
// 20 cutouts, one round trip, 160 creditsThe part where generation actually earns its keep
Normalising catalogue photos is editing work. The generation tools become interesting one step later, when you need the same product in six different settings — a summer campaign, a Christmas grid, a lifestyle shot for the category page.
A cutout plus edit_image gives you that without booking a studio for each one. The product stays the real photograph; only what surrounds it is generated.
"Place this chair in a sunlit Scandinavian living room, oak floor"
"Same product, autumn setting, warm low light, soft shadow"
"Studio shot on seamless grey, single soft key from the left"
Where to draw the line
This is worth being deliberate about, because the failure mode is expensive in returns and in regulatory attention.
The rule of thumb: the customer should be able to hold the item next to the picture and agree it is the same object.
Catalogue pipeline questions
Can I process a thousand product photos?
Yes, in batches. Group the calls into multicall requests so they execute concurrently rather than one at a time, and run them from a queue that records the resulting URL against each SKU. The credit cost is linear and predictable, which makes it easy to size before you start.
Is it legal to alter a supplier’s product photo?
That depends on your supplier agreement, not on the tool. Most distribution agreements permit resizing and background normalisation for retail listings, and many require it. Editing the product itself — changing its colour or shape — is a different question and one to ask your supplier before you automate it.
What about marketplace image requirements?
Most marketplaces specify a plain background, a minimum resolution and a maximum file size, with the product occupying most of the frame. That maps exactly onto this pipeline: cut out, place on white, upscale to clear the minimum, compress to clear the maximum.
Should I generate product photos rather than edit real ones?
For the product itself, no. A customer who receives something that does not match the picture will return it, and in many markets a materially inaccurate product image is a regulatory problem. Generate the surroundings — lifestyle settings, backdrops, seasonal scenes — and keep the product itself a real photograph.
What does one product cost to process?
Background removal 8, upscale 15, compression 1 — 24 credits for the full treatment, or 9 if the source is already high enough resolution. A thousand SKUs at the full treatment is 24,000 credits, so this is a job to size against a plan before you queue it.