Nano Banana API: Key, Pricing per Image and Free Tier (2026)

The Nano Banana API is Google's Gemini API, called with one of the Nano Banana image model IDs, such as gemini-nano-banana-2.1. You get a free API key in Google AI Studio, but image generation itself has no free tier: as of 31 October 2026, Google charges from $0.0336 per 1K image for Nano Banana 2.1 up to $0.24 per 4K image for Nano Banana Pro. If you'd rather not handle keys or billing, you can run the same models with credits on Kenerate AI or from an AI assistant through Kenerate's MCP server.

This guide covers the three things people search for most: how to get a Nano Banana API key, what each model costs per image, and what the free tier and rate limits really are. Every Google number below comes from Google's own pricing and developer pages, linked at the end. If you're new to the model itself, start with what Nano Banana is.

Is there an official Nano Banana API?

Yes, but it doesn't have "Nano Banana" in its name. "Nano Banana" is Google's nickname for its Gemini image models, so the official API is the Gemini API (for developers, via Google AI Studio) or Vertex AI (for Google Cloud and enterprise users). You pick the version by its model ID.

Many sites also sell "Nano Banana API" access as resellers. They wrap Google's model in their own API and set their own prices. That can be convenient, but the prices, uptime and data terms are theirs, not Google's.

How do you get a Nano Banana API key?

The key is a Gemini API key from Google AI Studio. Here are the steps from Google's API key docs (last updated 27 October 2026):

1. Open the API Keys page in Google AI Studio at aistudio.google.com/apikey and sign in with a Google account.
2. Accept the terms. New users get a default Google Cloud project and an API key automatically once they accept the Terms of Service.
3. Or import an existing project. If you already use Google Cloud, open Dashboard → Projects → Import projects, choose the project and import it, then create a key in it.
4. Check the key type. Google says new keys are auth keys by default. If a key shows as "Standard" and "Unrestricted", restrict it to the Gemini API or create a new auth key.
5. Set up billing for images. Every key belongs to a Google Cloud project, and that project's billing account pays for usage. Because image models have no free tier (see below), you need billing linked before image calls go through.
6. Store the key as an environment variable. Google's client libraries read GEMINI_API_KEY or GOOGLE_API_KEY. If both are set, GOOGLE_API_KEY wins.

How do you keep the key safe?

Google's own advice, in short:

• Treat the key like a password and never commit it to source control.
• Don't put it in a web page or mobile app. Call the API from your own server instead.
• Keep production keys in a secret store such as Google Cloud Secret Manager.
• Set billing alerts in the Google Cloud console and restrict keys by IP or origin.
• If a key leaks, create a new one, deploy it, then disable the old one.

What does a Nano Banana API call look like?

Google's current image-generation docs use the Interactions API. A minimal Python call with the official google-genai library looks like this (adapted from Google's sample):

pythonCopy
from google import genai
import base64

client = genai.Client()  # reads GEMINI_API_KEY from the environment

interaction = client.interactions.create(
    model="gemini-nano-banana-2.1",
    input="A studio product photo of a matte black coffee mug on a walnut table, soft morning light",
)

with open("mug.png", "wb") as f:
    f.write(base64.b64decode(interaction.output_image.data))

The REST equivalent is a POST to https://generativelanguage.googleapis.com/v1beta/interactions with your key in the x-goog-api-key header and a JSON body containing model and input. Google's docs also show optional settings for aspect ratio and image size (uppercase "1K", "2K", "4K"), Google Search grounding, thinking level, and previous_interaction_id for multi-turn edits. Check the docs for exact parameter names before you ship, because Google has changed the API surface more than once.

How much does the Nano Banana API cost per image?

Here are Google's list prices per output image, from the Gemini API pricing page (last updated 30 October 2026). "Standard" is a normal real-time call; "Batch" is the asynchronous Batch API.

Input is billed separately per million tokens (text, and images you upload as references). For Nano Banana 2.1 Google lists $1.50 per million input tokens on the standard tier. For a plain text prompt that's a small share of the cost; it matters more when you send several large reference images per call.

What does that mean in real money?

Some quick maths from the table, before input tokens:

• 1,000 Nano Banana 2.1 images at 1K: $33.60 standard, or $16.80 with Batch.
• 1,000 Nano Banana 2.1 images at 4K: $113 standard, or $56.70 with Batch.
• 1,000 Nano Banana Pro images at 2K: $134 standard, or $67 with Batch.

Three things follow. Batch halves the price if you don't need images back instantly. Pro costs the same at 1K and 2K, so 2K is the sensible default if you use it. And Nano Banana 2.1 is cheaper than Nano Banana 2 at every listed size while being the newer model, which is why Google points new projects at it. For a side-by-side of the two, see Nano Banana 2.1 vs Nano Banana 2, and for when Pro is worth it, Nano Banana vs Nano Banana Pro.

Is the Nano Banana API free?

No. Google's pricing page lists the free tier for every Nano Banana model as "Not available", so each API image is paid.

What is free:

• The API key itself. Creating a key in Google AI Studio costs nothing.
• Google AI Studio in the browser. The pricing page says "Google AI Studio usage is free of charge in all available regions." That's for trying models interactively, not for calling them from your own app.
• The Gemini app. Google's Gemini Apps limits page lists Nano Banana 2 image generation for users without a paid plan, at "standard limits". Again, that's the consumer app, not the API.

If you've seen "free Nano Banana API" offers, they're usually third-party resellers with trial credits. Read their terms carefully. For every free option we know of, see Is Nano Banana free?.

What are the Nano Banana API rate limits?

Google doesn't print image-model rate limits as fixed numbers on its rate-limits page (last updated 28 October 2026). It says image models are limited by images per minute (IPM) and that you can "view your active rate limits in AI Studio". Your limits depend on your usage tier:

Higher tiers get higher limits, and Google lists a spending cap per tier too. If you're planning a big batch job, check your project's limits in AI Studio first, and use the Batch API, which Google meters separately.

How do you use Nano Banana without an API key?

If you just want images, not a code integration, you don't need Google's API at all. Kenerate AI runs four Nano Banana models in its image studio, with no key, no Google Cloud project and no billing account. You pay in credits from a one-time pack.

Credit packs start at $15 for 1,300 credits, and credits never expire. Nano Banana models aren't on Kenerate's free plan: they unlock with any credit pack. New accounts get 25 free credits for the free-plan models (Qwen Image and Flux 2). Kenerate doesn't offer Nano Banana Pro.

To be straight about cost: per image, Google's own API is the cheaper route at volume. Kenerate makes sense when you don't want to write code, manage a key and a billing account, or juggle separate accounts for image and video models.

Can you call Nano Banana from Claude, ChatGPT or Cursor?

Yes, through the Kenerate MCP server. MCP lets an AI assistant use outside tools, so you can ask Claude, ChatGPT, Cursor, VS Code or Codex to "make a product shot with Nano Banana 2.1" and it generates the image on your Kenerate account.

1. Open kenerateai.com/app/mcp and pick your client.
2. Connect. Claude.ai and ChatGPT need no key: you sign in to Kenerate when you connect. Other clients can sign in or use a Kenerate API key created on that page.
3. Ask for an image and name the model (kenerate/nano-banana-2.1). Credits are spent at the same in-app prices.

Which route should you choose?

• You're building an app or a pipeline: the Gemini API with gemini-nano-banana-2.1, Batch where you can, and your key on a server.
• You need the best text and layouts: gemini-3-pro-image on the Gemini API. It isn't on Kenerate.
• You want images, not code: Kenerate's image studio, or the Gemini app for free light use.
• You work inside an AI assistant: the Kenerate MCP server.

Whichever you choose, tested prompts save money. Our Nano Banana prompts library shows real outputs before you spend anything.

Updated 31 October 2026. Google prices, free tiers and rate-limit rules are from Google's Gemini API pages on that date; Kenerate credit costs are from the app's model configuration. Both can change.