Nano Banana 2.1 for 37 images: Google's price vs Sume at 1K, 2K, 4K

A 37-image set on Nano Banana 2.1 costs $1.24 to $4.18 at Google's list and $3.70 to $7.40 on Sume. Per-tier arithmetic and what the gap buys you.

5 min readSume
All posts

A 37-image set on Nano Banana 2.1 costs $1.2432 (1K), $1.8648 (2K) or $4.181 (4K) at Google's published standard rate, and $3.70, $5.55 or $7.40 through Sume. Sume is the more expensive line for this one model, so the case for it has to be something other than the per-image price. This post shows the arithmetic so you can decide for your own volume.

The two price lists, side by side

Google lists Nano Banana 2.1 (gemini-nano-banana-2.1) at $30.00 per million image output tokens and states the per-image equivalents. A 1K image uses 1,120 tokens, a 2K image 1,680 and a 4K image 3,780. Sume's catalog lists nano-banana-2.1 as a fixed price per image. Multiplying each by 37 gives the set cost.

Per-image and 37-image prices, Google standard tier vs Sume list (Google read 2026-10-09, Sume catalog as of 2026-10-09)
SizeSume per imageSume x 37Google per imageGoogle x 37Sume minus Google
1K$0.1000$3.70$0.0336$1.2432$2.46
2K$0.1500$5.55$0.0504$1.8648$3.69
4K$0.2000$7.40$0.1130$4.1810$3.22

How big the gap is

The 4K row is the narrowest gap: $0.20 against $0.113 is a difference of $0.087 per image, or $3.219 across 37 images. The 1K row is the widest in relative terms: $0.10 against $0.0336.

I am not claiming Sume matches Google's rate. It does not, and the table says so. Google also publishes a lower batch tier on the same page, so the direct route is cheaper still for large offline runs.

What the extra money pays for

The case for routing through Sume is operational, not price. One workspace balance covers images, video, voice, music and render, generation jobs are durable and pollable at one job endpoint. If your pipeline already needs those, paying about $0.05 more per 2K image is a line item you can plan around. If you only call one image model in bulk, go direct.

Scaling the set up

The gap grows linearly. Ten sets of 37 images (370 images) at 2K cost 370 x $0.15 = $55.50 through Sume and 370 x $0.0504 = $18.65 at Google's standard rate. A thousand 4K images cost $200.00 against $113.00, a difference of $87.00.

A mixed set narrows it. Thirty 1K drafts and seven 4K finals cost 30 x $0.10 + 7 x $0.20 = $4.40 on Sume, and 30 x $0.0336 + 7 x $0.113 = $1.799 at Google. If your workload looks like this, the absolute difference is $2.60 per set of 37, which is usually smaller than the time spent wiring a second billing relationship.

Check the Google page again before you plan a quarter. Prices on a launch week can change, and the read date on this post is 2026-10-09.

A decision rule

Image quality, editing behavior and rate limits are outside this price comparison. Google's page also lists a free tier and other terms, and I have not compared them with Sume's plans.

  • Under 200 images a month on this model alone: the gap is under $20 at 2K ($0.0996 x 200 = $19.92), so convenience usually wins.
  • Thousands of images a month on this model alone: use Google directly and check its batch tier.
  • Mixed pipelines: price the whole pipeline, not one line. See the 100-video budget post for the video side.
  • Google prices come from the standard-tier line of its pricing page, read 2026-10-09.
  • Sume prices come from the public catalog at api.sume.com/v1/catalog, which needs no key.
  • Neither includes tax or any plan fees.

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume