Nano Banana: the tutorial to create professional visuals
(pricing, prompts, examples)

A visual for a post, a clean product photo, a banner for your website: these needs come back every week, and every time you have to choose between paying a designer, spending an hour in design software or publishing an average visual. Nano Banana, Google's image generator built into Gemini, changes that equation: you describe the visual in plain language and it produces it in seconds, readable text included. This Nano Banana tutorial explains which versions exist, what is free and what is paid, how to use it step by step and which requests to copy to get professional visuals today.

What is Nano Banana? The versions in September 2026

"Nano Banana" is the nickname Google gave to its family of image generation and editing models, built into the Gemini app. The name spread so widely that Google made it official. At the time of writing (September 2026), the family includes four models:

• Nano Banana 2 (Gemini 3.1 Flash Image), launched on February 26, 2026: the default model in the Gemini app, in Google Search and in Flow. Fast and versatile, it handles text in images, up to 5 consistent characters and output from 512 pixels up to 4K.
• Nano Banana Pro (Gemini 3 Pro Image), released in November 2025: the most precise model, built for complex compositions, multilingual text and brand consistency.
• Nano Banana 2 Lite, available since July 2026: the fastest and cheapest version, mainly aimed at developers generating visuals at high volume.
• Nano Banana (Gemini 2.5 Flash Image), the original August 2025 model, now considered a legacy version.

For an entrepreneur, the key point is simple: in the Gemini app, you use Nano Banana 2 without changing any setting, and you can switch to Nano Banana Pro when a visual needs more precision. To see where Nano Banana fits among other generators, our guide to creating images with AI covers the full landscape.

Is Nano Banana free or paid? Pricing

Nano Banana is free to try, with limits. Here are the benchmarks at the time of writing (September 2026), based on prices shown in France:

• No subscription: image generation and editing with Nano Banana 2 are included in the free Gemini app. Daily quotas vary with demand and are not guaranteed by Google.
• Google AI Plus (a few euros per month): higher limits, the option to redo a visual with Nano Banana Pro, and credits for Flow, Google's video tool.
• Google AI Pro, about €21.99 per month: much more comfortable quotas, 1,000 Flow credits per month and access to Gemini's advanced features.
• Google AI Ultra, from about €99.99 per month: the highest limits, only useful if images and video become the core of your business.

How much does Nano Banana Pro cost?

There is no separate "Nano Banana Pro" subscription. In the Gemini app, Nano Banana Pro is used through the "redo" option available from the Google AI Plus plan upward. For developers using the Gemini API, billing is per image: about $0.134 per 1K or 2K image and $0.24 per 4K image for Nano Banana Pro, compared with about $0.067 per 1K image for Nano Banana 2. Google states there is no free API tier for these models.

Our advice: start with the free version for one week, on your real needs. If you regularly hit the limits, Google AI Pro pays for itself as soon as it saves you one order from a designer. Always check current prices on Google's website, as they change often.

How to use Nano Banana: step-by-step tutorial

There is nothing to install. Here is the simplest method, in the Gemini app:

1. Open Gemini at gemini.google.com or in the mobile app, and sign in with a Google account.
2. Select the image creation tool in the input bar, or simply start your request with "Create a visual…". Gemini understands the intent.
3. Describe the visual with five elements: subject, setting, lighting, style, format (for example "square format" or "16:9 format").
4. Add references if needed: a photo of your product, your logo, your portrait. Nano Banana uses them to keep your product or your face accurate.
5. Edit through conversation: "make the background navy blue", "remove the mug on the left", "make the title bigger". No need to regenerate everything: the model only changes what you ask for.
6. Switch to Nano Banana Pro if the result lacks precision (long text, diagram, strict brand guidelines): use the option to redo the visual with the Pro model, if your plan includes it.
7. Download the visual and check it before publishing, especially the text.

There are two other entry points: Flow, Google's video tool, where Nano Banana 2 creates the starting images for your shots; and Google AI Studio, for developers who want to build Nano Banana into their own tools through the API.

Nano Banana in other languages: what to know

The Gemini app is available in many languages, and you can write your requests in your own language. Nano Banana 2 also displays readable text directly in the visual, accents included, and can translate the text of an existing visual into another language. This is very useful to adapt a poster or an ad for another market.

Two precautions. First, always put the text you want displayed in quotation marks in your request, so the model reproduces it word for word. Second, check every letter: on long texts, a typo or missing accent can still slip in. The shorter the text, the more reliable the result. To sharpen your wording, our beginner's guide to prompting gives you the basic method.

Nano Banana examples: 4 prompts for your business visuals

Here are four tested requests for the most common needs. Replace the elements in brackets with your own. Each one often saves one to two hours compared with a photo shoot or an outside order, depending on your situation.

1. A product photo ready for your online store

Add a photo of your product, even one taken with your phone, then:

"Using the attached photo, create a professional product photo: the [product] on a light marble countertop, soft natural light coming from the left, a subtle drop shadow, blurred beige background. Keep the exact shape, colors and label of the product. Square format, photorealistic."

2. A post visual for LinkedIn or Instagram

"Create a 4:5 post visual announcing my workshop. Plain navy blue background, large white title at the top: 'AI Workshop: 2 hours to save time'. Below, smaller: 'Thursday, October 16, 6 pm, online'. Clean style, modern sans-serif typography, plenty of empty space."

3. A website or profile banner

"Create a 16:9 banner for the website of a [business type] firm. Scene: a bright office with a wooden table, an open laptop and a plant, seen at a three-quarter angle. Leave the left third of the image empty and plain so I can place my text there. Main colors: [color 1] and [color 2]. Photorealistic."

4. A professional headshot from your photo

Add one or two sharp, well-lit photos of yourself, then:

"Using my attached photos, create a professional head-and-shoulders portrait: dark jacket, light shirt, plain gray background with a slight gradient, soft studio lighting, natural smile. Faithfully keep my facial features, hairstyle and skin tone. 1:1 format for a profile picture."

If the first result is not right, do not start over: say what is wrong ("the light is too harsh", "the product looks smaller than in reality"). That conversation is where the visual truly becomes yours.

Creative IA Pro

Produce all your visuals without a designer

In 3 days, Creative IA Pro has you build your branding kit, infographics, presentations and videos with AI. You leave with a method and deliverables ready to publish.

Discover Creative IA Pro

Nano Banana Pro vs Nano Banana 2: which one to choose?

Both models produce high-quality visuals. The difference comes down to precision and cost:

• Nano Banana 2 for the vast majority of your needs: post visuals, product photos, banners, quick edits. It is faster, free in Gemini and handles reference images very well.
• Nano Banana Pro when precision matters most: an infographic with a lot of text, a diagram with figures, a visual that must match brand guidelines exactly, text in several languages. It reasons more before producing the visual, which takes a little longer.

A simple method: always generate with Nano Banana 2 first. If the text or layout still does not hold after two or three adjustments, redo the visual with Nano Banana Pro.

Nano Banana vs Midjourney vs ChatGPT Images

Three tools lead visual creation in September 2026. Here is how they compare for professional use:

Nano Banana vs ChatGPT (GPT Image). Both are excellent for business visuals. OpenAI released ChatGPT Images 2.5 on September 8, 2026, with more precise editing and faster generation. If you already use ChatGPT every day, start there; if you work in the Google ecosystem, Nano Banana is the natural choice. The best approach is still to test the same request on both.

Nano Banana vs Midjourney. Midjourney remains ahead on artistic rendering (illustrations, moods, moodboards), but it is paid from day one and less comfortable editing an existing visual or placing precise text. For commercial visuals faithful to your product, Nano Banana is generally more effective. Our ChatGPT, Claude and Gemini comparison also helps you choose your main assistant.

Limits and mistakes to avoid

Publishing without proofreading the text. Even when well rendered, generated text can contain a typo. Check every word, figure and date.

Showing a product that does not exist. A visual can enhance, but it must not mislead: colors, size and finish must match what the customer will receive.

Using someone else's face. Do not create visuals from photos of real people without their consent. For your own portrait, there is no problem.

Forgetting that visuals are marked. Google embeds an invisible watermark (SynthID) and provenance metadata (C2PA) in generated visuals. Some platforms detect them and display an "AI content" notice. That is not a problem in itself, but it is worth knowing.

Sharing confidential documents. Avoid sending sensitive material (plans, client data) as reference images. For good practices, see our article on GDPR and AI.

FAQ: common questions about Nano Banana

Is Nano Banana free?

Yes, in the Gemini app, Nano Banana 2 can be used for free with daily limits. Nano Banana Pro and higher quotas require a Google AI subscription. The API is paid per image.

What is the difference between Nano Banana and Gemini?

Gemini is Google's assistant; Nano Banana is the model that creates and edits images inside Gemini. You go through Gemini to use Nano Banana.

Can Nano Banana visuals be used commercially?

Google allows you to use generated content, provided you follow its terms of use and respect third-party rights (brands, people, protected works). Review the current terms before any significant advertising use.

Can Nano Banana create videos?

No, Nano Banana creates images. For video, Google offers Veo and Gemini Omni. Our Veo 3, Kling and Sora comparison helps you choose.

How can I produce all my visuals faster?

By building your brand guidelines, request templates and formats once and for all. That is exactly what we do together in Creative IA Pro: branding kit, infographics, presentations and videos, in 3 days, without a designer.