Nano Banana 2.1 search grounding: 5,000 free, then $14 per 1,000
Nano Banana 2.1 grounding is $14 per 1,000 search requests after 5,000 free a month, shared across Gemini 3.x. Sume's image API has no grounding field.

Nano Banana 2.1 grounding costs nothing for the first 5,000 search requests a month and $14 per 1,000 after that, on Google's pricing page. The 5,000 is shared across all Gemini 3.x models on the project, so another model's traffic can use it up. The charge is for each search query the model makes, not for each API call.
How the bill adds up
Google's footnote says one request "may result in one or more queries to Google Search", and you pay for each query performed. Text and image results the model retrieves are not charged as input tokens. So the image price you saw earlier still applies and grounding is a separate line.
| Search queries in a month | Billable queries | Grounding cost |
|---|---|---|
| 5,000 | 0 | $0 |
| 20,000 | 15,000 | $210 |
| 100,000 | 95,000 | $1,330 |
What changed with 2.1
Google's image guide says Nano Banana 2.1 and Gemini 3.1 Flash Image add Image Search grounding next to web search, and that the Lite model does not support grounding at all. When you use Image Search, the guide says you must display the source attribution, and the API returns inline citations for grounded responses.
Turn grounding on only when the prompt depends on facts that change, such as a current product, a recent event or a landmark. A stylized poster of an invented object gains nothing from it.
Grounding through Sume
Sume's Image API docs list the parameters a model accepts, and a parameter that is not listed is refused with 400 unsupported_parameter. I found no search grounding field in the image parameters, so you cannot ask Sume's Nano Banana 2.1 to search. The usual workaround is to fetch the reference yourself and send it as an input_references image.
If grounded generation is the reason you chose the model, call Google directly for those requests.
Sources
Related posts
More in Pricing
- One 16:9 clip cropped to 9:16 and 1:1: $0.04, not 3 generations
A 15-second 720p Wan clip is $1.875. Cropping it to 9:16 and 1:1 with two $0.02 video-filter jobs adds $0.04, against $3.75 to generate both shapes again.
- What does one hour of AI narration cost on Sume, MAI and ElevenLabs?
One hour of finished narration is about 45,000 characters: $2.16 on Sume in three jobs, against $0.99 on MAI-Voice-2.1. Rates read 2026-10-07.
- One music bed for 50 holiday ads: $0.125 total, $0.0025 each
A Lyria music track is a flat $0.125 on Sume. One bed reused across 50 ads is $0.0025 per ad; a fresh track for every ad costs $6.25.
- Plan fee as a share of $1,000 usage: Pro 4%, Startup 12%, Scale 40%
At $1,000 a month of Sume generation, the plan fee is 4% on Pro ($40), 12% on Startup ($120) and 40% on Scale ($400). Usage bills separately.
Written by Sume