Amazon Image Compliance, Rule by Rule

Graflio does not have an automated compliance checker. As of August 2026 nothing in the app measures whether your background is pure white, whether the product fills enough of the frame, or whether a word has crept onto your main image. What it does is stay on the right side of the rules by construction — it composites your real product photograph and it builds gallery images, not main images — and leave the checking to you. This page is the checklist that replaces the button we do not have.

The rules that actually get listings suppressed

Amazon runs two rulebooks. The main image, the one that appears in search, is policed hard because uniform results are what makes the grid readable. Everything after it is far freer.

Note the two figures on the zoom row, because a lot of advice online collapses them into one. The stated floor to enable zoom is 1,000 px on the longest side. 1,600 px or more is what Amazon recommends. Anything you export from the Graflio Studio is 2160 px square and clears both; the file sitting in the gallery before you export is 1024 px, which passes the floor and misses the recommendation. If an image matters, export it rather than saving the gallery file.

What Graflio gets right without being asked

Three things are structural rather than checked. First, the product is your real photograph, composited into the scene — generation handles the background and the scene layers, not the product itself, so the most dangerous category of violation, an image of a product that is not the product, is not something the workflow produces on its own.

Second, the sets are gallery images by design: infographics, feature callouts, dimension diagrams, comparison panels, how-to sequences. There is no main-image mode, so nothing tempts you to upload a designed image into slot one.

Third, every image prompt ends with an explicit no watermarks instruction, and A+ modules are produced at Seller Central's module sizes, 970 × 600 for the standard modules and 970 × 300 for the banner.

That is the honest extent of it. None of the above is a check. If you upload a Graflio image as your main image, nothing stops you and Amazon will.

What you still have to verify yourself

The claims row is the one sellers underestimate. A research-driven brief writes confident headlines because it read your reviews and your competitors' reviews, and a headline like eliminates odour for 90 days is a claim you now own. Categories differ sharply here: supplements cannot make disease claims, beauty is constrained on before-and-after, and anything implying a health outcome deserves a second read. The tool will not stop you. A reviewer might.

A five-minute pre-upload pass

Do this once per SKU, in this order.

One: open the main image, sample the corners, confirm 255,255,255 and not 254,254,254 or a warm off-white that looks white on your monitor. Two: check the product's longest dimension against the frame and crop if it is swimming in space. Three: confirm slot one has no text, badge, logo, border or prop on it. Four: check the resolution of every file you are about to upload, and re-export the ones under 1,600 px. Five: shrink the whole set to thumbnail size and read it — most infographics that fail do so on mobile, not on policy. Six: read the headlines as claims rather than as copy, and soften anything you cannot evidence.

If you want an actual audit rather than a checklist, that market exists: seller suites with listing quality scoring flag weak or non-compliant imagery as part of a broader listing score. SellerApp's listing quality checker is the clearest example — it names the problem, it does not fix it. Pairing a diagnostic tool with a production tool is a perfectly sensible stack, and we would rather say that than pretend we ship both.

Where Graflio helps with compliance, and where it does not

The editable text point is the one that pays off after a rejection. If a reviewer objects to one word, a flat AI image means regenerating and re-checking everything; a text object means retyping the word and exporting again. Compliance work is iterative, and the cost of iteration is the thing worth optimising.

Common questions

Does Graflio check my images against Amazon's rules before upload?

No. There is no automated compliance check in the product as of August 2026: nothing measures background whiteness, frame fill, resolution or text on the main image. The workflow avoids the structural violations — your real photo is composited, and the output is gallery imagery rather than main images — but the pre-upload check is manual.

What resolution should an Amazon listing image be?

At least 1,000 px on the longest side to enable zoom, with 1,600 px or more recommended by Amazon. Graflio's Studio exports at 2160 px square, which clears both. The generated file in the gallery is 1024 px, so export rather than saving that file when the image matters.

Can I use a Graflio infographic as my main image?

No. The main image must show only the product on a pure white background, with no text, logos, watermarks, borders or props, and with the product filling roughly 85% of the frame. Infographics belong in every slot after the first.

Are AI-generated listing images allowed on Amazon?

The line Amazon polices is misleading imagery, not the tool that made the image. Graflio composites your real product photograph and uses generation for scenes and backgrounds, which keeps the product itself accurate. A generated product, a generated size, or a generated feature the item does not have is where policy and returns both go wrong.

What happens if my main image breaks the rules?

Amazon can suppress the listing from search until the image is fixed, which is invisible in your sales reports until you notice traffic has stopped. Fix the cause — background, frame fill, text, resolution or prohibited content — and re-upload the corrected file.