Nano Banana Pro vs GPT Image 2.5 for 9:16 covers: 18.8 vs 3.6 cents

For a 9:16 cover, Nano Banana Pro bills $0.1875 per image on Sume and ChatGPT Image 2.5 at high quality bills about $0.0356 at 720x1280. Twenty covers are $3.75 against $0.71. Both list 9:16, so the ratio is not the deciding factor; price, resolution control and your own eyes on the output are. The Image API docs define both request shapes.

This is a cost comparison, not a quality ranking. Vendor claims about how each model looks belong on each vendor's own page, and the only honest test is the same prompt on both.

How do I ask each model for 9:16?

Nano Banana Pro takes a ratio and a resolution tier. GPT Image 2.5 takes either a ratio or explicit pixels, and explicit pixels are what the price below assumes. When both image_size and aspect_ratio are sent, image_size has priority.

What do twenty covers cost?

Nano Banana Pro's row is the 1K default; its 4K tier is listed higher and Sume scales the reservation for it. GPT Image 2.5 moves with quality.

How do I decide?

Generate three covers per model for the same prompt, put them side by side, and look for text errors and edge artifacts at phone size. If GPT Image 2.5 is good enough, it is 5 times cheaper at high. If you need Nano Banana Pro's resolution tiers or 21:9 and 4:5 ratios, that is what you pay for. Prices here are list x 1.25 in dollars per image before whole-cent rounding; the catalog's billable formula reads "list x 1.25, ceil usd cents", so confirm the first charge in usage.cost and budget from the invoice, not from the sum.