250 SKUs, two 4K heroes each: $100.00 or $111.25 on Sume

The short answer

Two 4K hero images for each of 250 SKUs is 500 images. On Sume that costs $100.00 with Nano Banana 2.1 (500 x $0.20) and $111.25 with GPT Image 2.5 high (500 x $0.2225). At 2K the totals are $75.00 and $27.81.

At 4K the two models are close in price, while at 2K GPT Image 2.5 medium is far cheaper.

The arithmetic

The 4K tier gap between the two is $0.0225 per image, so for 500 images the difference is $11.25. At 2K the gap is $0.094375 per image, or $47.19 for 500.

Google lists $0.113 for a 4K image, so 500 would be $56.50 direct, and $28.35 at its batch rate.

How to run it

Do a 10-SKU pilot at the target tier first. That is 20 images, $4.00 on Nano Banana 2.1 at 4K, and it surfaces prompt problems before they cost $100.

Plan for 202 envelopes. At 4K and high quality the generation may not finish in the 30-second budget, so poll the job rather than assuming a body.

Retries and cost per SKU

Some images will be rejected by eye, and rejected images are billed like any other, so budget a margin. The bullets add 20 percent to each Sume total in the table above (the totals multiplied by 1.2) and divide by the number of SKUs to give a per-SKU cost.

Treat the per-SKU figure as the number to compare across models, because it does not change when the count changes. Re-read the public catalog before a large run, since prices can change and the table is a snapshot from 2026-10-08.

• Nano Banana 2.1 at 4K: $100.00 becomes $120.00 with 20 percent extra images, which is $0.48 per SKU (a 250-SKU job).
• GPT Image 2.5 high (4K): $111.25 becomes $133.50 with 20 percent extra images, which is $0.53 per SKU (a 250-SKU job).
• Nano Banana 2.1 at 2K: $75.00 becomes $90.00 with 20 percent extra images, which is $0.36 per SKU (a 250-SKU job).
• GPT Image 2.5 medium (2K): $27.81 becomes $33.37 with 20 percent extra images, which is $0.13 per SKU (a 250-SKU job).

What the docs say to check

Check the status code, not the body shape. POST /v1/images waits up to 30 seconds and returns the images with 200. Slow settings (4K, high quality, a large n) are the most likely to come back as 202 with a job envelope, and you then read the images from the job result endpoint described in Jobs and results.

On GPT Image 2.5 the quality values are auto, low, medium, high, xhigh and max. If you omit quality the default is high, and auto reserves max, so pin the tier you budgeted for instead of leaving it to the default.

If hero images need only web size, 2K halves the price on GPT Image 2.5 and cuts it by a quarter on Nano Banana 2.1.

The numbers above are list prices read on 2026-10-08, not measurements. Sume figures are the catalog prices that already include its margin, Google figures are from its own pricing page, and quality or speed differences are not covered. Re-read both pages before you commit a large budget, because either side can change its rates.