Image 2.5
AI Image Model

Nano Banana Pro

Studio-grade images from a model that reasons first

Generate and edit with Google’s Gemini 3 Pro Image: accurate infographics, legible multilingual text, multi-image blending, studio lighting and camera control, and 2K or 4K output.

Creative workspace

Nano Banana Pro

27

What is Nano Banana Pro?

Nano Banana Pro is Google DeepMind's Gemini 3 Pro Image model. It reasons about a request before it draws, which is why it handles infographics, legible text in several languages, and edits that combine many reference images while keeping people recognisable.

How the people who built the model describe it:

“Built on Gemini 3 Pro, Nano Banana Pro uses Gemini's state-of-the-art reasoning and real-world knowledge to visualize information better than ever before.”
Google

Nano Banana Pro at a glance

Model
Gemini 3 Pro Image
References
Up to 8 here
Output
1K / 2K / 4K

What you can make with the Nano Banana Pro tool

These are the jobs people actually bring to Nano Banana Pro, and the settings each one needs.

01

Explainers and infographics

Turn a recipe, a process, or a set of notes into a diagram whose labels and steps are actually right.

02

Localised campaign assets

Keep the design and swap the language on packaging, posters, and signage for another market.

03

Product and design comps

Combine a sketch, a material, and a mood reference into one finished render.

Nano Banana Pro

A workspace shaped for the model

Nano Banana Pro puts the handful of controls that change the result most in front of you, and keeps the rest out of the way.

Visuals that use real-world knowledgeOfficial model example · Google DeepMind

01

Visuals that use real-world knowledge

Google built it on Gemini 3 Pro so it can reason about the subject: ask for a how-to infographic and the steps, quantities, and labels come out in the right order rather than as decorative filler.

Legible text, in more than one languageOfficial model example · Google DeepMind

02

Legible text, in more than one language

Headlines, signage, and lettering render as real, correctly spelled words in varied fonts and textures — including when the text is built out of the scene itself.

Translate and localise a finished designOfficial model example · Google DeepMind

03

Translate and localise a finished design

Give it an existing image and a target language: it rewrites the text in place and keeps the layout, lighting, and product intact.

Blend references into one designOfficial model example · Google DeepMind

04

Blend references into one design

Google says the model can combine up to 14 input images and keep up to five people consistent. On this site you can attach up to eight references per request.

Studio control over light, focus, and angleOfficial model example · Google DeepMind

05

Studio control over light, focus, and angle

Relight a scene from day to dusk, move the focus, change the camera angle or aspect ratio, and finish at 2K or 4K without starting over.

Nano Banana Pro

Choose with context

Where Nano Banana Pro is the right call, and where a different model does the job better.

Compared with Nano Banana 2

More finish and fidelity; less focused on speed

Compared with GPT Image 2

A strong fit for visual finish and reference-led realism

Best decision signal

The asset needs a premium 2K or 4K final

From brief to studio in three steps

01

Set the creative anchor

Write the idea and add any source or reference material the tool needs.

02

Shape the output

Choose the model, format, resolution, and duration that match delivery.

03

Continue in Studio

Review the complete draft in the shared conversation flow. Generation will be enabled when the model API is connected.

Nano Banana Pro FAQ

Start Creating With Nano Banana Pro

Set up the brief now, then continue in the shared Studio workspace.