Nano Banana 2.1 is Google’s newest AI image model, and it started rolling out on October 6, 2026 in the Gemini app, AI Mode in Search and several other Google tools. The big improvements are cleaner text inside images, support for up to 14 reference photos, and people and products that stay consistent from one image to the next. Here’s how to use it, prompts that get better results, how Google labels AI images, and what to watch out for before you use them for your business.
Key takeaways
- Rolling out since October 6, 2026 in the Gemini app, AI Mode in Search, Google AI Studio, Flow, Stitch, Google Ads and Gemini Enterprise.
- Better text rendering, so signs, menus and infographics are more readable.
- Blend up to 14 reference images; keep up to 4 characters and 10 objects consistent.
- Output at 1K, 2K or 4K, including very wide and tall formats.
- Images carry Google’s invisible SynthID watermark. Label AI images honestly when you use them commercially.
What’s new in Nano Banana 2.1
Nano Banana is Google’s nickname for its Gemini image models. Version 2.1 is an update to Nano Banana 2 (also known as Gemini 3.1 Flash Image), designed to be faster and cheaper to run than Google’s top-end Nano Banana Pro. According to Google’s developer documentation and coverage from Android Authority, the main changes are:
- Cleaner text in images. Better spelling and layout for posters, menus, labels and infographics.
- Up to 14 reference images. Upload photos of a product, a person, a room and a style, and combine them in one result.
- Consistent characters. Up to 4 people or characters and up to 10 objects can stay recognizable across a series of edits.
- Higher quality at 1K, 2K and 4K, plus a fix for the repeating “tiling” glitches that used to appear in very wide panoramas.
- Search grounding. It can use Google Search to get real-world details right, which helps with infographics and diagrams.
- Better editing, including changing just one area of an image without redoing the rest.
How to use Nano Banana 2.1 in the Gemini app
- Open the Gemini app (or gemini.google.com) and sign in with your Google account.
- Tap the tools menu and choose Create images (it has a banana icon).
- Pick a mode if you’re offered one: Fast, Thinking or Pro.
- Upload reference photos if you want to edit or combine images. You can add several.
- Write your prompt (tips below) and send.
- Refine in conversation. Ask for one change at a time: “make the background a brick wall,” “move the logo to the top left.” You don’t need to start over.
- Download the version you like.
Free or paid? Google’s Gemini image page lists the free tier as using Nano Banana 2, and says Nano Banana 2.1 is unlocked with a paid Google AI plan, with higher limits on paid plans. Availability can vary by country and device, so check what your account shows. In AI Mode, you can ask Search to create or edit an image right in the results, where AI Mode is available.
Prompts that work better
Google’s own advice is to describe the action, subject, setting and mood, then add details about composition, style and aspect ratio. Here are prompt patterns built on that advice for everyday and small business uses:
- Product shot: “Using my product photo, place the candle on a marble kitchen counter in soft morning light, shallow depth of field, square 1:1, no extra text.”
- Menu or flyer: “Create a 4:5 poster for a Brampton bakery’s Diwali special. Headline exactly: ‘Sweet Box Special – $24.99’. Warm gold and deep red colours, clean modern font, lots of empty space.”
- Consistent character: “Using the person in photo 1, show them in three scenes: opening a shop door, serving a customer, and waving goodbye. Keep the same face, hair and green apron in each.”
- Room or storefront mockup: “Using photo 1 (my storefront) and photo 2 (my logo), show the logo as a lit sign above the door at dusk. Keep the building exactly the same.”
- Infographic: “Make a simple 9:16 infographic with 4 steps to winterize a car in Ontario. Large readable text, icons, high contrast.”
Tips: put any text you need in quotation marks and say “exactly.” Number your reference photos in the prompt (“photo 1,” “photo 2”). Always proofread text in the result. It’s better than before, but no AI model gets spelling right every time.
SynthID, labelling and commercial use
Google says images made in the Gemini app carry an invisible SynthID watermark plus a visible mark showing they’re AI-generated. You can upload an image to the Gemini app and ask whether it was made with Google AI. That’s useful when a “photo” you see online looks too perfect. (We explain how that check works, and its limits, in our SynthID Detector guide.)
Before you use Nano Banana images for your business, keep these cautions in mind:
- Don’t mislead customers. An AI image that makes a product, a renovation or a food item look better than the real thing can cross into misleading advertising, which Canada’s Competition Act prohibits. Use AI for concepts, backgrounds and mockups, not as “proof” of real work.
- Get permission for people’s photos. Uploading a customer’s or employee’s face to make new images of them needs their consent. Never create images of real people in situations they didn’t agree to.
- Avoid other people’s brands and art. Don’t upload logos, characters or artwork you don’t own and then sell the result.
- Ownership is still a grey area. Google has said in its terms that it won’t claim ownership of content you generate, but whether AI images can be protected by copyright in Canada is still unsettled. Don’t build your main logo on a purely AI-generated image.
- Label it. Platforms like Instagram, Facebook and YouTube have their own AI-content labels. Using them builds trust.
Station take
Nano Banana 2.1 is a meaningful upgrade for anyone who needs quick visuals with readable text, like menus, social posts, flyers and mockups. The 14-reference limit and consistent characters make it practical for small brands that want a series of images that look like they belong together. But it’s a tool for drafts and concepts, not a replacement for real photos of real work, and anything you publish should be checked by a human. For more tools like this, see our 8 free AI tools for small business owners, and for how Google’s AI is changing Search, read Google starts paying publishers for AI Overviews. Follow the latest in Trends.
FAQ
Is Nano Banana 2.1 free?
Google’s Gemini page lists Nano Banana 2 as the free model and Nano Banana 2.1 as part of its paid Google AI plans in the Gemini app. Limits and availability vary by region, so check your own account.
How many reference images can Nano Banana 2.1 use?
Up to 14 reference images, with up to 4 consistent characters and up to 10 objects, according to Google’s documentation.
Can people tell my image was made with AI?
Yes, often. Google embeds an invisible SynthID watermark, and the Gemini app can check for it. Gemini app images also carry a visible AI mark.
Can I use Nano Banana images in ads?
Generally yes, as long as you follow Google’s usage policies, don’t use people or brands without permission, and don’t mislead customers about what they’re buying. Nano Banana 2.1 is also built into Google Ads for creating ad images.
Sources: Google Gemini image generation · Google on Threads · Android Authority · Let’s Data Science · 9to5Google. This article was prepared with AI assistance and reviewed by Media Station. Disclosure.