Nano Banana Pro at 2K ($0.1875) vs Nano Banana 2 at 4K ($0.20) on Sume

On Sume, a Nano Banana Pro image at 1K or 2K costs $0.1875, a Nano Banana 2 image at 4K costs $0.20, and a Pro image at 4K costs $0.375. If you want the most pixels for the least money, Nano Banana 2 at 4K is the pick; if you want Pro's output at normal sizes, 2K is the same price as 1K.

These are Sume list prices times 1.25, from the pricing tables in the repo: Pro lists $0.15 at 1K and 2K and $0.30 at 4K, and Nano Banana 2 lists $0.06, $0.08, $0.12 and $0.16 at 0.5K, 1K, 2K and 4K.

All the tiers

Google's own lists, for scale

Google lists Gemini 3 Pro Image at $0.134 for 1K or 2K and $0.24 for 4K, and Gemini 3.1 Flash Image at $0.067 (1K), $0.101 (2K) and $0.151 (4K) (Gemini pricing, read 2026-10-06). The ordering is the same on both: Pro is dearer than Flash at equal size, and Pro's 4K step is the big one.

Sume's prices are not Google's prices, because Sume's provider rates and its 1.25 factor differ; use the Sume table for a Sume budget.

Picking one

Pro has a flatter 1K-to-2K price, so there is no saving in dropping to 1K. Nano Banana 2 has a 0.5K tier and ratios up to 8:1, which Pro does not list. Pro's catalog list of ratios is 10 plus auto, and Nano Banana 2's is 14 plus auto.

The honest tradeoff is quality you can see against price you can budget. Render the same prompt on both at the size you will ship before you decide, using the comparison post as a starting point.

Check it on your own account

Do not budget from a blog table alone. GET /v1/images/models lists every model with its descriptors, and GET /v1/images/models/{id}/endpoints shows the pricing line for one model. Then run one small request and read usage.cost on the response, which is the billed amount in USD; the token counts in usage are reported as 0 on this route.

Run the test at the quality and size you plan to ship, because both move the price. A single test at low quality costs under a cent for most sizes here, so it is a cheap way to confirm your assumptions before a batch.

Sync, async and failures

The /v1/images route waits up to 30 seconds for the image. If the job finishes in that window you get the result directly; otherwise you get a 202 and an async job to poll. Write your client to branch on the status code, since larger sizes and higher quality are the likely cases for a 202.

Requests are strict. A parameter the chosen model does not list returns 400 unsupported_parameter, stream returns a 400, and provider.only or provider.order accept only sume. Treat a 400 as a bug in the request, not a transient error, and do not retry it unchanged.