Nano Banana Pro costs 1.25x the 2.1 at 2K on Sume, 4x at Google 1K
Google lists Pro at $0.134 and 2.1 at $0.0336 for 1K (about 4x). Sume bills Pro $0.1875 and 2.1 $0.15 at 2K, only 1.25x. The ratio at each size.

The short answer
Nano Banana Pro costs 1.25 times Nano Banana 2.1 at 2K on Sume: $0.1875 against $0.15. At Google's own list prices the same two models are far further apart at 2K, $0.134 for Pro against $0.0504 for 2.1, a ratio of 2.66. So the price gap you read on Google's page is not the gap you pay on Sume, and it changes by size.
The ratios, size by size
Google's figures come from its Gemini API pricing page (read 2026-10-09): Pro at $0.134 per 1K or 2K image and $0.24 per 4K image, and 2.1 at $0.0336, $0.0504 and $0.113 for 1K, 2K and 4K. Sume's figures are billed per image to your wallet.
| Size | Google 2.1 | Google Pro | Google ratio | Sume 2.1 | Sume Pro | Sume ratio |
|---|---|---|---|---|---|---|
| 1K | $0.0336 | $0.134 | 3.99x | $0.10 | $0.1875 | 1.875x |
| 2K | $0.0504 | $0.134 | 2.66x | $0.15 | $0.1875 | 1.25x |
| 4K | $0.113 | $0.24 | 2.12x | $0.20 | $0.375 | 1.875x |
Where the two ladders bend
On Sume, Pro has one price for 1K and 2K ($0.1875) and doubles to $0.375 at 4K. Nano Banana 2.1 climbs steadily: $0.10, $0.15, $0.20. So the cheapest place to buy Pro on Sume is 2K, where it costs only $0.0375 more than 2.1 ($0.1875 minus $0.15). At 1K the extra is $0.0875. At 4K it is $0.175.
For 100 images at 2K, that is $15.00 for 2.1 and $18.75 for Pro, a difference of $3.75.
When Pro is worth it
Pro is the choice when the 2.1 output misses on a hard prompt, such as dense text or a complex scene. If 2.1 works, the 25% saving at 2K applies to every image. Try 2.1 first, and move only the failed prompts to Pro.
The two models share the tier names, so switching in code means changing the model id and nothing else.
What to check before you switch
Price is only one input. The models take different reference limits and aspect lists: the repo's catalog gives Pro an 11-value list (with auto) and 2.1 a 15-value list that adds 4:1, 1:4, 8:1 and 1:8. If you need an extreme banner, 2.1 is the only one of the two that lists it. Read each model's supported_parameters from GET /v1/images/models and test one prompt on both before you commit a batch.
For a worked example, take 200 images at 2K. Nano Banana 2.1 costs 200 x 0.15 = $30.00, and Nano Banana Pro costs 200 x 0.1875 = $37.50. The $7.50 difference is the price of choosing Pro for the whole set. If only 20 prompts need Pro, the mixed bill is 180 x 0.15 + 20 x 0.1875 = $27.00 + $3.75 = $30.75.
At 4K the gap widens: 2.1 is $0.20 and Pro is $0.375, a ratio of 1.875. That is a reason to check whether you need 4K at all before choosing Pro.
A last point: the Google figures in this post come from Google's own pricing page, read 2026-10-09, and apply to Google's API. Sume's prices are its own list prices and include Sume's markup, so the two ratios are not meant to be compared as a statement about quality.
Sources
Related posts
More in Pricing
- 18,000 characters: Sume TTS $0.85 vs ElevenLabs $0.72–$1.44
Voicing 200 IVR prompts of 90 characters: Sume $0.0475 per 1,000 characters is $0.85; ElevenLabs lists $0.04 Flash, $0.08 multilingual. Read 2026-10-09.
- 60,000 characters: Sume TTS $2.85 vs ElevenLabs $2.40–$4.80
Voicing 500 social captions of 120 characters: Sume $0.0475 per 1,000 characters is $2.85; ElevenLabs lists $0.04 Flash, $0.08 multilingual. Read 2026-10-09.
- 1,000,000 characters: Sume TTS $47.50 vs ElevenLabs $40.00–$80.00
Voicing a 1,000,000-character library: Sume $0.0475 per 1,000 characters is $47.50; ElevenLabs lists $0.04 Flash, $0.08 multilingual. Read 2026-10-09.
- 10,000 characters: Sume TTS $0.47 vs ElevenLabs $0.40–$0.80
Voicing a 10,000-character blog post narration: Sume $0.0475 per 1,000 characters is $0.47; ElevenLabs lists $0.04 Flash, $0.08 multilingual. Read 2026-10-09.
Written by Sume