How to Fix Product Images to Meet Amazon's Resolution Requirements

I helped a friend list his first batch of products on Amazon last year. He had about 60 SKUs, mostly kitchen tools. Nice photos, shot on a decent camera. We uploaded everything, filled out the listings, and went to bed thinking we were done. Woke up to 42 rejection emails. Almost all of them said the same thing: images did not meet technical requirements.

Some were too small. Some had the wrong background. Some had borders or watermarks or text overlays that Amazon's system flagged. We spent the next three days fixing every single one. Here is what Amazon actually checks and how to make sure your images pass on the first try.

The exact resolution requirements and why they exist

Amazon requires your main product image to be at least 1000 pixels on the longest side. That is the hard floor. Below 1000 pixels, you get an automatic rejection. But 1000 pixels is actually the bare minimum. To enable the zoom function on your listing, which lets customers hover to see details, you need at least 1600 pixels on the longest side. No zoom means fewer conversions. Every product where we used exactly 1000 pixels got listed but sold less than the ones where we pushed to 1600.

There is also a lower bound. Images with any side below 500 pixels get rejected outright. We had a few products where the photo was 1200 by 400 after cropping, and every single one was rejected. The system does not care that the long side meets the requirement. If any dimension drops below 500, it fails.

The background has to be pure white. Pure white means RGB 255, 255, 255. Not off white. Not cream. Not light gray. Amazon's system actually checks this programmatically. About a third of our rejections were for backgrounds that looked white to our eyes but measured at something like 250, 250, 250 on the color picker. Close enough is not close enough.

How we fixed the low resolution photos

About 25 of the rejected photos were simply too small. My friend had cropped them aggressively to remove backgrounds, and the resulting images fell below 1000 pixels. The original uncropped photos were fine, around 3000 pixels wide. But we had deleted the originals because we thought the cropped versions looked better.

Lesson one. Always keep your original photos. But we did not have them, so we had to fix the cropped versions. We tried just scaling them up in Photoshop first. The 800 pixel images became 1600 pixels but looked noticeably soft. Photoshop's standard resize just stretches pixels and fills in the gaps with blurry approximations. Amazon did not reject them, but the zoom function looked terrible. Customers hovering over a garlic press saw a blurry mess instead of crisp details.

Then we tried AI upscaling. The difference was immediate. A 2x upscale on those 800 pixel images produced clean 1600 pixel versions that looked sharp enough for Amazon's zoom feature. The AI model generated new pixel detail based on what it understood the product to look like, rather than just stretching the existing pixels. The garlic press looked like a garlic press again.

Processing cost was 2 credits per image on ClarifyPix. For 25 images, that was 50 credits total. About the cost of a sandwich, and it saved us from having to re-shoot 25 products. Each upscale took about five seconds.

Fixing the white background in bulk

About a dozen rejections were for non-white backgrounds. These were photos shot on what my friend thought was a white surface. It was a light gray cutting board. In person it looked white enough, but the camera picked up the gray, and Amazon's automated check flagged it.

We ran these through AI background removal to strip the existing background and replace it with pure white. On ClarifyPix, background removal costs 1 credit per image. For twelve images, that was 12 credits and about 30 seconds of processing. The tool output each image with an RGB 255,255,255 background. Every single one passed Amazon's check on the first re-submission.

One thing I learned: after background removal, check the edges at 200% zoom. AI removal on simple products like kitchen tools is very clean. But sometimes you get a faint halo where the old background met the product. On a white replacement background, this halo is invisible to the naked eye but it can show up in Amazon's automated check. A quick pass with a soft eraser at the edges fixes it in seconds.

The other requirements that got us rejected

Amazon is picky about more than just resolution and background. Here is what else they flagged in our batch.

No text or graphics on the image. We had a few photos where the product packaging was visible, with brand names and descriptions printed on it. Those got rejected. The main image must show only the product itself, not its packaging with text.

No borders. One photo had a thin white border from a previous editing step. It was maybe three pixels wide. Amazon caught it. Trim your images to the exact edge of the canvas.

The product must fill at least 85% of the image frame. We had a few shots where the garlic press was small in the center with lots of white space around it. Those were rejected. Crop tighter. The product should dominate the frame.

A checklist that saves hours of re-submission

After fixing those 42 rejections, I made a checklist. Now I run through it before uploading anything to Amazon. It takes about 30 seconds per image and catches everything that would trigger a rejection.

Check one: longest side at least 1600 pixels, shortest side at least 500. Check two: background is RGB 255,255,255 when sampled with the color picker. Check three: no text, no logos, no watermarks, no borders visible anywhere in the frame. Check four: product fills at least 85% of the image area. Check five: image is in focus at 100% zoom.

If any of these fail, fix the image before uploading. AI upscaling handles check one. AI background removal handles check two. Cropping handles check four. The rest are things you catch by looking carefully at each image. Five minutes of checking upfront saves hours of rejection emails and re-uploading.

Amazon's requirements are strict but they are also predictable. Once you know exactly what the system is checking for, compliance is mechanical. The hard part is not meeting the requirements. It is knowing what they are in the first place. Now you know.