Nano Banana Pro Guide: Free Limits, API Pricing and How to Use It

Nano Banana Pro is Google’s top image model, the one people use for posters with readable text, product shots, infographics and edits that keep a face looking like the same person. It is also confusingly sold in five different places, each with its own limits.

This guide clears that up: what Nano Banana Pro is, how it differs from Nano Banana 2, every free and paid way to use it, daily limits, API prices per image, step-by-step use in Gemini, Flow and AI Studio, prompt tips, watermarks, and fixes for the usual errors.

What Is Nano Banana Pro?

“Nano Banana” is Google’s nickname for its Gemini image models. Nano Banana Pro is the high-end one, officially Gemini 3 Pro Image, built on Gemini 3 Pro’s reasoning.

Google pitched it at work where small details matter:

• Readable text: posters, menus, infographics and diagrams with correct spelling, including other languages.
• Consistency: keeping people, products and styles the same across edits and multiple reference images.
• High resolution: 1K, 2K and 4K output.
• Precise edits: change lighting, camera angle, focus or one object while keeping the rest.
• Grounded facts: it can use Google Search to make infographics or diagrams about real things.

Nano Banana Pro vs Nano Banana 2 vs the Original

Google Flow also lists newer variants such as Nano Banana 2.1 and Nano Banana 2 Lite. Prices are from Google’s Gemini API pricing page, checked 6 November 2026.

Is Nano Banana Pro Free?

Partly. There are three genuinely free routes, and none is unlimited:

• Google Flow (labs.google): up to 20 Nano Banana Pro images a day without a paid plan. The cap resets 24 hours after your first image, and failed or blocked images do not count.
• Gemini app, free tier: creates images with Nano Banana 2, not Pro, under “standard limits” that refresh every few hours up to a weekly cap.
• Google AI Studio: lets you try models in the browser, but the Pro image model has no free API tier.

Warning: any site promising “free unlimited Nano Banana Pro” is a third party reselling the API or using another model. Many ask for payment or your Google login. Stick to Google’s own domains.

Limits by Plan

Google does not publish exact image counts for the Gemini app and adjusts them with demand. Full plan comparisons are in my Google AI Plus vs Pro vs Ultra guide.

API Pricing per Image

Under the hood Google bills image output at $120 per million tokens and input at $2 per million, so long prompts and many reference images add a little. Since 1K and 2K cost the same, always ask for 2K unless you need 4K.

Tip: generate drafts with Nano Banana 2 ($0.067 at 1K), then run only the keepers through Pro at 2K or 4K. More on every Gemini price in my Gemini API pricing guide.

How to Use It in the Gemini App

1. Open gemini.google.com or the Gemini app and sign in.
2. Choose image creation from the tools menu (the banana icon or “Create image”), or simply ask “create an image of…”.
3. Write your prompt or upload one or more reference photos to edit.
4. Switch to Pro: on a paid plan, pick the thinking/Pro model before you send, or use Redo with Pro on a Nano Banana 2 result.
5. Refine in chat: “make the title bigger”, “change the background to sunset”, “keep everything else the same”.
6. Download the full-size image from the image menu.

How to Use It Free in Google Flow

1. Go to labs.google/flow and sign in with your Google account.
2. Create a project and choose image generation.
3. Select Nano Banana Pro as the model.
4. Prompt and generate. Image generation does not use Flow’s video credits.
5. Track your count. Free accounts get 20 Pro images per 24 hours from the first image.

Note: in India, South Korea and Vietnam, Flow adds a visible watermark automatically unless you are on the top Ultra tier.

How to Use It in AI Studio and the API

1. Open aistudio.google.com, pick Gemini 3 Pro Image and test prompts in the playground.
2. Turn on billing for your project; the Pro image model is not on the free tier.
3. Create an API key under Get API key.
4. Call the model from code (Python example below).

pip install google-genai

from google import genai
from google.genai import types

client = genai.Client()  # reads GEMINI_API_KEY
resp = client.models.generate_content(
    model="gemini-3-pro-image",
    contents="A clean infographic titled 'How Solar Panels Work', 4 labelled steps, flat style",
    config=types.GenerateContentConfig(
        response_modalities=["IMAGE"],
        image_config=types.ImageConfig(aspect_ratio="16:9", image_size="2K"),
    ),
)
for part in resp.candidates[0].content.parts:
    if part.inline_data:
        open("infographic.png", "wb").write(part.inline_data.data)

Check Google’s image generation docs for the current parameter names in your SDK version, since config fields are updated between releases.

Prompt Tips and 8 Example Prompts

• Describe the scene, not keywords: subject, setting, lighting, camera, style, mood.
• Put exact text in quotes: a sign that reads "OPEN 24 HOURS".
• State the format: aspect ratio, resolution, “poster”, “product photo”, “isometric diagram”.
• For edits, say what must not change: “keep her face, hair and outfit identical”.
• Iterate in small steps rather than rewriting the whole prompt.

1. “A minimalist coffee shop menu board with 6 drinks and prices, chalk style, text perfectly legible.”
2. “Product photo of a white sneaker on a pastel podium, soft studio light, 4K, shallow depth of field.”
3. “An infographic explaining how much VRAM AI models need, 4 bars, labelled 8 GB, 16 GB, 24 GB, 32 GB.”
4. “Turn this photo into a 1950s travel poster with the words ‘VISIT GOA’ at the top.” (upload photo)
5. “Change the lighting in this room to golden hour and keep all furniture the same.” (upload photo)
6. “A YouTube thumbnail background: neon city at night, empty space on the left for text.”
7. “A labelled diagram of a computer’s insides for a school project, simple and colourful.”
8. “Combine these three product photos into one flat-lay on a marble table.” (upload photos)

Watermarks and SynthID

• Invisible SynthID: every Nano Banana image carries Google’s hidden SynthID watermark.
• Visible mark: Google said Gemini app images for free and AI Pro users carry a small visible sparkle, removed for Ultra and AI Studio.
• Checking an image: you can upload an image to the Gemini app and ask whether it was made with Google AI; it checks for SynthID.

Warning: removing watermarks or using the model to make misleading images of real people breaks Google’s policies. Label AI images when you publish them.

Nano Banana Pro vs GPT Image vs Open Models

Read ChatGPT Images 2.6 for OpenAI’s side, or the best open-source text-to-image models if you want unlimited images on your own GPU.

Troubleshooting

I cannot find Nano Banana Pro in the Gemini app

Free accounts get Nano Banana 2. Upgrade to Google AI Plus or higher, or use Flow for 20 free Pro images a day.

“You’ve reached your limit” or images stop generating

You hit the daily or weekly cap. Wait for the reset (Flow resets 24 hours after your first image), switch to Nano Banana 2, or upgrade.

“I can’t create that image” or safety block

The prompt touched a blocked area (real people in sensitive situations, explicit content, logos used deceptively). Rephrase without real names or sensitive details; blocked tries in Flow do not count toward your 20.

Text in the image is misspelled

Make sure Pro is selected, put the exact words in quotes, keep text short, and ask for “large, clear lettering”. Fix single words with a follow-up edit.

The face changes between edits

Upload the original as a reference each time and say “keep the person’s face exactly the same”. Start a new chat if the conversation is long.

API error: model not available on free tier

Enable billing on your Google Cloud or AI Studio project; gemini-3-pro-image has no free quota.

My Take

Nano Banana Pro is the image model I reach for when the result has to be right the first time: thumbnails with text, simple infographics, product mock-ups. The 20 free images a day in Flow are plenty for most hobbyists.

For volume or private work, I still generate locally with open models in ComfyUI, which cost nothing per image once you have the GPU. If you are new to that, start with my VRAM guide.

FAQ

Is Nano Banana Pro free?

Partly. Google Flow gives up to 20 free Nano Banana Pro images a day. The Gemini app’s free tier uses Nano Banana 2, and the API has no free tier for Pro.

What is the difference between Nano Banana Pro and Nano Banana 2?

Pro (Gemini 3 Pro Image) is better at text, detail and consistency. Nano Banana 2 (Gemini 3.1 Flash Image) is faster and about half the price.

How much does Nano Banana Pro cost per image?

$0.134 per 1K or 2K image and $0.24 per 4K image on the API, or half that with Batch.

Can Nano Banana Pro make 4K images?

Yes, through the API and AI Studio. In Flow, 4K upscaling is limited to Ultra.

Do Nano Banana images have a watermark?

All have an invisible SynthID watermark. Free and AI Pro Gemini app images can also have a small visible mark.

Is nanobanana.com an official Google site?

No. Google’s official places are gemini.google.com, labs.google/flow and aistudio.google.com.

Can I use Nano Banana Pro images commercially?

Google’s terms generally let you use what you create, but you must follow its usage policies and respect other people’s rights. Check the current terms for your account type.

Sources

• Google: Gemini API pricing (image models)
• Google blog: Introducing Nano Banana Pro
• Aliteq: Is Nano Banana Pro free? Gemini and Flow limits (updated 3 Nov 2026)
• Google: Gemini API image generation docs