Models

Images (Optional)

Prompt
0/4096

Not sure what to write? Try an example:

Nano Banana 2.1

Upload multiple reference photos of people, outfits, or products, specify each role, and instantly generate a unified high-resolution image with consistent details and legible text, up to 4K.

An indoor snapshot of a woman in a white T-shirt peeling away like a banana skin to reveal her in a yellow linen suit at a sunny harbor, next to a banana crate stenciled NANO BANANA 2.1.

What Is Gemini Nano Banana 2.1?

Listed in Google's developer docs as Gemini Nano Banana 2.1, this is the image generation and editing model Google released on October 6, 2026 as an update to Nano Banana 2. Google positions it as the more efficient counterpart to Nano Banana Pro, with better image quality, clearer text and steadier character consistency than the version before it. Its strength is working from several reference photos at once: hand it a person, an outfit, a product and a location in separate images, and it returns one picture that keeps all of them recognizable. On this page you can write a prompt on its own, or upload your photos and describe how they fit together, then download the result.

A product photo of an olive waxed canvas backpack and a photo of a mountain trail viewpoint combined by Nano Banana 2.1 into one shot of the backpack on the granite rock, with the Tancky AI logo and the site address.
What's new

What's New in Nano Banana 2.1

Combining several photos used to go wrong in familiar ways: a face that drifts after the second edit, a logo that comes back misspelled, a product that changes shape. This update works on exactly those weak spots.

Up to ten reference photos

Upload as many as ten photos in one request and give each a role in the prompt, such as image 1 for the person and image 2 for the jacket.

Faces and products that stay the same

People, characters and products keep their features from one scene to the next, so a series of images still shows the same person or the same item.

Cleaner text and infographics

Headlines, labels and short captions come out more legible, and posters or explainer graphics hold their layout better. Put the exact words in quotes for the best results.

Sharper detail in wide shots

Overall image fidelity is significantly sharper, delivering seamless transitions and crisp panoramas in 2K and 4K without edge artifacts.

What you can make

Image Editing With Multiple Reference Photos

Three jobs that depend on the model reading several photos at once.

A portrait of a man in a gray t-shirt and a cream cable-knit sweater combined into one image of the same man wearing the sweater.

Put an outfit on a person

Upload a portrait and a product photo of a jacket, dress or pair of sneakers. The person keeps their face and pose, and the clothing keeps its color, pattern and logo, which helps you preview a look before you shoot it.

A fox mascot reference image next to three scenes of the same fox with a green scarf: a kitchen, a park and a bedroom.

Keep one character across scenes

Start from one image of a character or mascot, then ask for new settings one at a time. Each result keeps the same face, outfit and proportions, so the set works as a storyboard, a comic or a series of social posts.

A water bottle photo, a logo and a beach photo combined into an ad of the bottle on the sand with the logo and the headline Stay Fresh.

Turn a product and a logo into an ad

Combine a product shot, your logo and a photo of the setting you want. The result places the product in that setting, puts the logo where you asked and adds a short headline in clean lettering.

Three steps

How to Use Nano Banana 2.1 Online

  1. A corgi photo and a yellow dog raincoat photo in the upload area, numbered 1 and 2.
    Step 01

    1. Upload your photos

    Add a photo of your dog and a product photo of a yellow raincoat. The order matters: the first photo becomes image 1, the second image 2, and so on.

  2. The two thumbnails above a prompt asking to put the raincoat from image 2 on the dog in image 1, with 1K selected.
    Step 02

    2. Say what each photo adds

    Write that the raincoat from image 2 goes on the dog in image 1, then describe the scene, such as a rainy city sidewalk. Pick a resolution: 1K is enough to check the idea, 2K or 4K for the version you keep.

  3. The corgi wearing the yellow raincoat on a wet city sidewalk, with a download button in the corner.
    Step 03

    3. Compare and download

    Check that the dog's face and the raincoat's color and strap match your photos. If something drifted, name it in the prompt and run it again, then download the version you like.

Who it helps

Who Gets the Most From Multi-Photo Editing

Anyone who already has the right photos and needs them combined into a new image.

Online sellers

Show one product in several settings, or put clothing on a model, without booking a new shoot for every listing.

Storytellers and creators

Keep a recurring character recognizable across storyboards, comics, thumbnails and posts.

Marketing teams

Turn a product shot, a logo and a mood photo into campaign drafts in different formats for review.

Multi-Image AI Synthesis · Fast Generation

Combine Your Photos With Nano Banana 2.1

Upload reference photos of people, products, and scenes, describe your vision, and generate a cohesive high-resolution image with consistent features.

Nano Banana 2.1 FAQ

Answers about what changed, how it compares with other Nano Banana versions, reference photos, credits and commercial use.