A rejected main image is one of the few Amazon listing errors you can catch before submitting — pure white background, 85% frame fill, 1000px+ for zoom, no text or watermarks. Yet most sellers only learn the rules after a rejection email.
So I opened every free compliance checker I could find and tested the same things on each:
- Does the page actually have a working upload? (I checked the rendered HTML for
type="file", not just the marketing copy) - Does the image stay on your device, or is it uploaded to a server?
- What exactly gets checked?
- Is there a limit, login, or upsell?
- Can you fix the image afterward, or only see the error?
Disclosure up front, not buried at the end: I work on PicViva, one of the tools in this list. I've tried to test ours as harshly as the others, and I'll point out where competitors beat us — because in two cases, they do.
The comparison table
| Tool | Runs locally | Checks | Login | Fix it after? | Languages |
|---|---|---|---|---|---|
| Fernpix | Yes — "on-device, nothing uploaded" | White bg, 85% fill, zoom size, format, watermark, centering, edge-crop | No | Separate fix flow | EN |
| zonfy | Yes (pixel checks); AI check uploads | White bg, fill, resolution, shape | No | Yes — separate Image Fixer + bulk processor | EN |
| Listimo | Pixel checks local; "KI-Blick" is AI | Size + rules, plus AI detection of text/logos/watermarks | No | Within listing suite | DE |
| zproductai | Yes — "runs in your browser" | Pure white RGB 255, fill, zoom size, file type | No | No (editor is a separate paid product) | EN |
| Toololis | Client-side (upload appears after JS) | 12 rules: 85% fill, RGB 255 bg, 1000px+, no logos, no lifestyle on main | No | No | EN |
| ReddStudio | Couldn't verify | Markets a checker; the page I loaded (8.6 KB) showed no upload before JS | No | Part of paid AI image service | EN |
| PicViva (ours) | Check runs in editor, free tier | Size, background, text, watermark, product coverage | No | Yes — same tool, zero extra steps | 19 |
Now the details, in the order I tested them.
Fernpix — the most complete rule coverage
The strongest pure checker I found. Three upload slots, and it checks things others skip: centering, edge-crop risk, watermark detection, and it separates main image rules from secondary image rules — which matters because Amazon allows lifestyle shots in slots 2–9 but rejects them in slot 1. It also gates checks by destination marketplace, including eBay, Etsy and Lazada variants. Everything runs on-device.
Where it's weak: it's a checker plus a separate fix flow. You still take the error list somewhere else to actually repair the image.
zonfy — the checker that wants to be embedded
zonfy's checker does the pixel rules locally and clearly labels what's local vs. what requires their AI. But the interesting part is the ecosystem around it: a dedicated Image Fixer (square crop, ~85% fill, true #FFFFFF, 2000px export) and a bulk processor for whole listings. They even offer the checker as an embeddable widget for seller guides — a smart distribution play I haven't seen elsewhere.
One honest note: their fixer docs describe "the near-white trap" — backgrounds that look white but measure 240-something. That's exactly the failure mode I wrote about in my last post: a global brightness boost can't fix it without clipping your product. Good to see others hitting the same wall.
Listimo — the one that proved me wrong
I assumed non-English markets had no local-language tooling. Listimo is fully German: the checker page, the rules, the suite around it (a 0–100 listing score, a "Rufus check" that tests whether your listing answers buyer questions). And they do something none of the English tools do well: they split the problem honestly. Pixel-measurable rules run locally; text, logos and watermarks get an AI pass, because no pixel measurement catches a "FREE SHIPPING" badge.
That split — "what code can measure" vs. "what needs a model" — is the most honest architecture in this list.
zproductai — minimal, clean, no tricks
Does exactly what it says: pure white (RGB 255), fill, zoom dimensions, file type. In-browser, no login, no upsell mid-check. The fewest checks here, but zero friction. Its sibling compliance-claims-checker (for listing claims) is a nice touch.
Toololis — a checklist that grades itself
12 rules with pass/fail output, including "no lifestyle imagery on main" — a rule most pixel tools skip because detecting a lifestyle scene needs more than thresholds. The upload widget renders after JavaScript loads, so my static HTML scan found no file input; judge it in your browser, not by my curl.
ReddStudio — listed for completeness
Their sitemap advertises an image checker, but the page I loaded showed no upload UI before JS, and their core product is paid AI listing images. I can't verify what the checker does, so I won't claim it works.
PicViva (ours) — where we win, and where we lose
Win: the check and the fix are one flow. The checker opens directly in the editor (I verified GET /en/app/editor/new?tool=amazon-image-compliance-checker returns 200 with no login redirect), flags size, background, text, watermark and coverage issues, and the toolbar to fix each one is right there — resize, background removal, the compliance re-check. And the tool page exists in 19 languages, because marketplace sellers overwhelmingly aren't browsing in English.
Lose, honestly: zonfy ships 13 free seller tools and a bulk processor; we ship one standalone free tool. Fernpix covers more marketplaces. Listimo's AI text/logo pass catches things our pixel checks don't. If you need any of those today, use theirs — this post will still be here when you need a fixer.
Which one should you actually use?
- Privacy-first, widest rules: Fernpix
- Whole-catalog cleanup in one go: zonfy's fixer + bulk processor
- German-language workflow: Listimo
- One image, zero friction: zproductai
- Check → fix → re-check without switching tools: PicViva — that single-flow design is the entire reason ours exists
Every tool above is free at the point of check. None of them will get your image un-rejected after the fact — test before you upload, not after Amazon emails you.
Method note: all findings collected Nov 2026 by loading each page and inspecting the rendered HTML; "runs locally" claims are taken from each tool's own documentation and spot-checked where feasible. Tools change — if one of these now behaves differently, that's worth knowing too.
