Amazon product image requirements in 2026

Most sellers treat the main image as a design decision. Amazon treats it as a compliance requirement. That gap is the most common reason a new listing never appears in search — not the keywords, not the price, the photo.

A main image with a prop, a border, a watermark, or a background that is nearly white can get a listing suppressed. Suppressed means it exists, it is live in your inventory, and it does not show up. Amazon does not always tell you why.

Here is what the rules actually are, and which of them are hard requirements versus guidelines that only affect conversion.

The main image: hard requirements

These are not stylistic preferences. Breaking them risks suppression.

| Requirement | Detail | | --- | --- | | Pure white background | RGB 255, 255, 255. Not off-white, not cream, not a gradient | | Product only | No props, no accessories that are not in the box, no packaging | | No added graphics | No text, logos, watermarks, borders, badges or inset images | | Product fills the frame | At least 85% of the image area | | Real product | Photograph or photorealistic render — no sketches or placeholders |

"Nearly white" is the one that catches people. A studio shot on a white sweep often comes out at RGB 250-252 in the corners, which reads as grey against Amazon's own white page and fails the check. It has to be pure.

The 1,600px threshold

Amazon enables zoom on images that are at least 1,600 pixels on the longest side. Below that, no zoom.

This is not a compliance rule — a smaller image will not get you suppressed. But zoom measurably lifts conversion, because shoppers who cannot inspect a product are more likely to leave than to guess. There is no reason to sit under it: the cost of a larger file is nothing compared to the cost of a shopper bouncing.

Upload larger where you can. Amazon downsamples for display and keeps the original for zoom.

The other image slots

The main image carries the compliance burden. The rest are where you actually sell, and the rules relax considerably — you can use lifestyle scenes, text overlays, infographics and comparison charts.

Most listings use two or three slots and leave the rest empty. That is the cheapest conversion win available on the platform. A full set typically covers:

• Scale. Something that shows real size. Buyers return items that arrive smaller than they pictured, and a returned item costs far more than the photograph would have.
• In use. The product in the context it belongs in.
• Detail. Material, texture, finish, closures — whatever the description claims, shown.
• What is in the box. Every component laid out. This pre-empts the single most common negative review: "I thought X was included."
• Specification callouts. Dimensions and key features as an infographic.

If you sell apparel or wearables, a shot on a person is close to mandatory — buyers cannot judge drape or fit from a flat lay.

Common failures, in order of how often they happen

1. Off-white background. Covered above. Check the corner pixels, not your impression of the image.

2. Text in the main image. A "50% OFF" badge or a brand banner in the main image is a clear violation. Text is fine in secondary images.

3. Props that are not included. A mug photographed with coffee beans scattered around it, where you are not selling the beans. This is a violation in the main image and a returns problem everywhere else.

4. Product too small in the frame. Under 85% and it can be rejected. It also looks worse in a search grid where your competitors fill their thumbnails.

5. Reusing the same photo across slots. Six near-identical angles is a wasted opportunity, not a complete gallery.

Does this apply to other marketplaces?

No, and assuming it does is its own mistake. Every marketplace has its own image rules:

• Walmart also wants a white background but applies different size and count requirements.
• Etsy actively rewards lifestyle and process imagery — a pure white packshot often underperforms there.
• eBay is far more permissive, and used-condition listings benefit from honest photos of actual wear.
• TikTok Shop is video-first; a still gallery is the secondary asset, not the primary one.

Shooting one set and pushing it everywhere is how sellers end up compliant on one channel and underperforming on three.

What good looks like

We publish a page of real generated output you can open at full size and inspect. Two things are worth checking on any AI-generated product image, ours or anyone else's:

Does the text survive zoom? Most image models produce text that looks plausible at thumbnail size and falls apart when you open it — misspelt words, invented letterforms, broken kerning. On an infographic that is the difference between an asset you can publish and one you redo by hand.

Do the proportions hold? Scale drift is the other tell. Generic models stretch a strap, thicken a sole, or size a mug like a bucket. A shopper who feels misled by the photo returns the item and says so in a review.

Generating the set

The AI product image generator produces six image types from a single product photograph: white background, lifestyle, infographic, lifestyle context, use case and promo banner. The white-background type is built to main-image rules specifically — pure white, product filling the frame, no props or text.

Each image costs 2 credits. Background removal is bundled at no extra cost. The free plan includes 10 credits a month with no card, which is enough to generate a full set for one product and judge the output on something you actually sell.

One thing to know before you start: generated output is deleted automatically by a daily purge, typically within about 48 hours, on every plan. Download what you want to keep as soon as it is generated.

---

Before you publish anything, check your listing text against the character limits for each marketplace with the free listing character counter, and work out whether the product survives Amazon's fees with the FBA fee calculator. A compliant image on a product with no margin is still a product with no margin.