Amazon A+ comparison carousel images: 488x700 portrait, 600x450

The Premium Comparison Carousel needs 3 to 5 product images of at least 488x700 plus 600x450 mobile images. Sume-legal sizes are 800x1152 and 960x720.

4 min readSume
All posts

The PremiumComparisonCarouselModule needs a portrait product image of at least 488 by 700 pixels per product, and a mobile image of at least 600 by 450. Do not simply round to 496 by 704 on Sume: that is only 349,184 pixels, below the 655,360 minimum for custom sizes. Ask for 800 by 1152 portrait and 960 by 720 mobile instead.

Amazon's numbers are from its A+ Premium modules page; Sume's rules are from the image generation docs, read 2026-10-01.

What does the comparison carousel require?

Amazon lists 3 to 5 products, each with a productTitle of 1 to 25 characters, a 10-character ASIN, and 3 to 7 comparison metric rows. The images are the part you generate.

Comparison carousel image minimums and Sume-legal sizes, read 2026-10-01.
ImageAmazon minimumAsk Sume forPixels
desktopImage (portrait)488 x 700800 x 1152921,600
mobileImage600 x 450960 x 720691,200
Too small to request488 x 700 rounded to 496 x 704Not valid349,184

Which size rules apply?

Both edges must be multiples of 16, maximum edge 3840, aspect ratio at most 3:1, and 655,360 to 8,294,400 pixels in total. Documented for openai/gpt-image-2.5 and openai/gpt-image-2.5-sunburst.

800 and 1152 are both multiples of 16 (50 and 72 times 16), and the ratio is 1.44:1, well inside the 3:1 limit.

How many calls for five products?

Five products need five portrait images and five mobile images. The example model descriptor in the docs shows n as a range of 1 to 4, but that is per model, so read the model's own limits. Use one request per product and orientation so each prompt names its product.

How do I keep the products consistent?

Use the same prompt skeleton, background and lighting words for every product, and change only the product description. Check each result against the minimums before you assemble the module.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume