Nano Banana API pricing: cost per image by resolution
Current Nano Banana API pricing for Nano Banana 2, Pro and 2 Lite: flat per-image and token billing, cost by resolution, and when each billing fits.
Read as MarkdownNano Banana API pricing on SeedRouter comes in two forms. For Nano Banana 2 and Nano Banana Pro, Standard model IDs charge one flat price per delivered image, at any resolution, and Official model IDs bill the tokens each render reports, so the price rises with the image size. Nano Banana 2 Lite is sold on Official token billing only. Failed requests are not charged.
The tables below are read live from the rate table that bills your account, for Nano Banana 2, Nano Banana Pro and Nano Banana 2 Lite. Nothing here is a remembered quote.
What does one Nano Banana image cost right now?
Nano Banana 2
Current rates are temporarily unavailable. No price estimate is provided; unavailable rates must not be interpreted as free usage.
Nano Banana Pro
Current rates are temporarily unavailable. No price estimate is provided; unavailable rates must not be interpreted as free usage.
Nano Banana 2 Lite
Current rates are temporarily unavailable. No price estimate is provided; unavailable rates must not be interpreted as free usage.
Each model has two tables. The first is the rate itself: a price per delivered image on Standard, and prices per million tokens on Official; Nano Banana 2 Lite shows its Official rate only. The second converts them into the cost of one image at every size the model supports. If a rate cannot be loaded, the table says so instead of showing zero; treat it as unknown, never as free.
Why does resolution change the price on Official?
Because each size produces a fixed number of image tokens, and Official bills those tokens. Google publishes the counts:
| Image size | Nano Banana 2 | Nano Banana Pro | Nano Banana 2 Lite |
|---|---|---|---|
512 | 747 | not offered | not offered |
1K | 1,120 | 1,120 | 1,120 |
2K | 1,680 | 1,120 | not offered |
4K | 2,520 | 2,000 | not offered |
Two details matter. On Nano Banana Pro, 1K and 2K produce the same number of image tokens, so a 2K image costs the same as a 1K one. On Nano Banana 2, a 4K image uses 2.25 times the image tokens of a 1K image. The aspect ratio does not change the count: a 16:9 1K image uses the same tokens as a square one.
On Standard, none of this applies. A 4K image costs the same as a 512 draft.
How do you estimate an Official request?
estimated USD = (
input tokens × input rate
+ image output tokens × image output rate
+ text and thinking output tokens × text output rate
) / 1,000,000Image output is the large term. Text and thinking output is small but not zero: measured on short 1K prompts, Nano Banana 2 reported 405 such tokens, Nano Banana Pro 246 and Nano Banana 2 Lite 398. Those are single samples; a longer prompt or reference images add input tokens, and reasoning varies with the prompt. The calculators on the Nano Banana 2, Nano Banana Pro and Nano Banana 2 Lite pages apply the same formula to the size you pick.
Standard or Official: which billing fits?
| Workload | Better fit | Why |
|---|---|---|
| 2K and 4K finals | Usually Standard | The flat price does not rise with size |
| A fixed budget per image | Standard | Cost is the image count times one price |
| Nano Banana 2 Lite | Official | Lite is sold on token billing only |
| Long reasoning on hard prompts | Standard | Thinking tokens are billed on Official only |
Compare the two columns of the live table for your model and size before a large run. Both IDs of a model accept the same request body, so the choice can change per request.
What does Google charge for the same models?
Google's Gemini API pricing page bills the same models per token and converts each size into a per-image equivalent, counting image output only. It uses the same image token counts as the table above, so the sizes that cost more on Google cost more on Official here too. The same page lists no free tier for any of the three models on the Gemini API.
When is a request not charged?
A task that ends failed is not charged, and neither is one that returns no image. A request blocked by the safety filters ends failed. A polling timeout on your side is not a failure: keep the task ID and check the same task again, because submitting a second request means paying for a second image. The task guide explains the difference.
Frequently asked questions
What is the cheapest Nano Banana model?
Compare the live tables above. Nano Banana 2 Lite has the lowest token rates of the three on Official, but it renders at 1K only. For large images, the flat Standard price of Nano Banana 2 or Nano Banana Pro can cost less than a token-billed 4K render.
Is Nano Banana Pro more expensive than Nano Banana 2?
On Official, yes: its image, text and input token rates are all higher. On Standard, check the live table; the flat price is set per model.
Where can I see what a request cost?
In usage history. The task response reports token usage when it is available, but it carries no dollar amount.
Price it before you send it
Pick the channel per job: a flat price for large finals, token billing when small drafts come in cheaper. Check the live tables above before a big run, top up once on the billing page with credits that never expire, and use the Nano Banana 2 API reference for the exact request fields.
