Google's Nano Banana Tutorial: Your First 10 Minutes

The first time I heard "Nano Banana" I assumed it was a meme, not a Google product. Turns out it's the cheeky nickname for Gemini's built-in image model, and honestly it's one of the more fun AI things I've played with this year. This Google's Nano Banana tutorial is the version I wish I'd had on day one: no fluff, just what to click and type in your first ten minutes so you actually make something instead of staring at a blank box.

Quick context before we start. "Nano Banana" covers Gemini's native image generation. There's a fast everyday version (Nano Banana 2, based on Gemini 3.1 Flash Image) and a fancier one called Nano Banana Pro (Gemini 3 Pro Image) that does sharper text and higher resolutions. You don't need to memorize any of that. You just need the Gemini app.

Minute 1-2: Open Gemini and Find Image Mode

Head to the Gemini app (web or mobile) and sign in with your Google account. You can start on the free tier, though free users only get a few low-resolution generations before you'd need a paid plan for the full 2K and 4K stuff.

To turn on images, open the tools menu and pick the "Create images" option (it literally has a little banana on it, which is how I knew I was in the right place). From the model menu you can choose Fast, Thinking, or Pro. For your first try, Fast is perfect. It's quick and totally good enough to learn on.

Minute 3-5: Write Your First Prompt

Here's where most people freeze, so let me make it easy. A good prompt for Nano Banana usually names a few things: the subject, what it's doing, where it is, and the style. You don't need fancy art words. Plain English works.

Try copying one of these to get going:

• Simple: "A cozy coffee shop on a rainy afternoon, warm lighting, photo style."
• With a subject: "A small orange cat sitting on a stack of books by a window, soft morning light, illustration."
• Product-ish: "A glass jar of honey on a wooden table, clean studio background, bright and simple."

Type it, hit send, and wait a few seconds. That's your first image. It probably won't be perfect, and that's completely normal. The real magic with this tool isn't the first try, it's the editing, which is the next thing I want you to learn.

Minute 6-7: Edit With Words (This Is the Good Part)

This is what makes Nano Banana actually worth your time. Instead of starting over, you just talk to it. After it makes an image, type a follow-up like you're chatting with a designer friend.

Real examples that work well:

• "Make it nighttime instead."
• "Change the cat to a dog, keep everything else the same."
• "Add a steaming cup of coffee on the table."
• "Remove the text in the corner."

It keeps the parts you liked and changes only what you asked for. You can also upload your own photo and tell it to edit that, which is honestly the feature I use most. "Here's my product, put it on a marble countertop with soft light" gives you a usable shot in seconds. The same trick works for headshots, room photos, and old pictures you want to brighten or clean up.

One more thing that surprised me: you can hand it more than one image at once and ask it to combine them. Nano Banana lets you mix a stack of reference images in a single prompt, so you can say "put the jacket from the first photo onto the person in the second photo" and it'll do a decent job. That's also how you keep a character looking the same across several images, which is great if you're making a little story or a set of matching social posts.

Minute 8: Pick the Right Shape and Size

By default you might get a square, but you usually want something else. Nano Banana supports loads of aspect ratios, including 1:1, 4:3, 16:9 (widescreen), and 9:16 (tall, like a phone screen or social story).

The easy trick: just describe the shape in your prompt. Say "vertical 9:16 social post" or "widescreen 16:9 banner" and it'll usually listen. If you want crisper, bigger files, look for the resolution dropdown to push it up to 2K or 4K, though that's where the paid plan comes in.

Minute 9: Try Text in an Image

One thing this Google's Nano Banana tutorial would be incomplete without: text. Most image AIs are notoriously bad at spelling, but Nano Banana Pro is genuinely good at putting readable words into an image, like a poster or a mock label.

A tip straight from Google's own guide that actually helps: chat with Gemini first to nail down the wording, then ask it to make the image with that exact text. So you might say, "Help me write a short slogan for a bakery," agree on one, then say "Now make a poster with that slogan." Doing it in two steps gives you cleaner text than cramming everything into one prompt.

A Few Prompt Recipes Worth Stealing

Once you've got the basics, the difference between a meh image and a good one is usually just how you phrase things. After a lot of trial and error, here are the patterns I reach for again and again. Steal them and swap in your own subjects.

• For a clean product shot: "[your item] on a [surface], soft studio lighting, plain [color] background, sharp focus, no clutter."
• For a lifestyle vibe: "[your item or person] in a [real setting], natural window light, candid feel, shallow depth of field."
• For an illustration: "[subject] in a flat illustration style, simple shapes, [two or three colors], friendly and modern."
• For a poster: "A poster for [thing], bold title that reads '[your words]', clean layout, lots of empty space."

The pattern is always the same: name the subject, then the lighting, then the style, then any text. When something looks off, change one piece at a time rather than rewriting the whole prompt. That way you actually learn what each word is doing.

Beginner Mistakes I Want You to Skip

I made all of these so you don't have to. None of them are dramatic, but fixing them early saves a lot of frustration in your first week.

• Cramming everything into one giant prompt. It's tempting to describe twelve things at once, but you'll get a muddle. Start simple, then refine with follow-ups.
• Giving up after the first result. The first image is a draft. The good stuff comes from two or three rounds of editing with words.
• Forgetting to say the shape. If you don't mention an aspect ratio, you'll often get a square and wonder why it doesn't fit your banner.
• Trusting fine details. Check text, hands, logos, and small print before you use an image anywhere real. The model still slips up on those.

Minute 10: Save It and Know the Catch

Download your image (there's a save or download button right on it) and you're done. Congratulations, you made something.

Now the honest caveats, because I won't pretend it's flawless. The free tier runs out fast, so if you fall in love with it you'll bump into limits or need to pay for the higher-res outputs. It can still mangle hands, fine details, and complicated scenes. And like all these tools, it sometimes ignores part of your prompt and you just have to ask again. None of that is a dealbreaker. It just means your first results are a starting point, not the finish line.

The Bottom Line

Ten minutes is genuinely enough to go from "what is a Nano Banana" to making and editing real images. Open Gemini, turn on image mode, write a plain-English prompt, then refine it by chatting. That edit-with-words flow is the whole reason this tool is fun, so spend most of your time there. Start on the free tier, see if you actually keep reaching for it, and only pay once you know you want the bigger, sharper outputs. For a first session, you really can't break anything, so go make something a little silly and learn as you go.