PCMag editors select and review products independently. If you buy through affiliate links, we may earn commissions, which help support our testing.

Google Unleashes Gemini 3 on New Nano Banana Pro AI Image Generator

Nano Banana Pro can generate text in multiple languages and handle up 14 images per prompt.

 & James Peckham Reporter

Our team tests, rates, and reviews more than 1,500 products each year to help you make better buying decisions and get more from technology.

Our Expert
LOOK INSIDE PC LABS HOW WE TEST
65 EXPERTS
43 YEARS
41,500+ REVIEWS
(Credit: Google)

Google's next-gen Nano Banana image-generation tool is here, days after the launch of its Gemini 3 model. Nano Banana Pro, also known as Gemini 3 Pro Image, promises more factually accurate creations. Google says it's "more helpful" than the previous version, and suggests using it to build explainers, infographics, diagrams, and blueprints.

An example shows the AI making a step-by-step recipe for a cardamom tea—researching the process, designing an image with text instructions, and organizing it into a flow chart structure.

An AI-generated image for a recipe
(Credit: Google)

AI tools have struggled with generating text in images, but Google thinks it has cracked the case with Nano Banana Pro, touting its ability to do everything from short taglines to long paragraphs. An example shows how text can be used in an AI-generated storyboard, putting multiple instances of small directions next to sketches.

(Credit: Google)

In a separate tips and tricks post, Google says the tool can be limited when trying to render small text, and it warns that spelling can be inaccurate. You may need to tweak images with follow-up prompts to get accurate results.

Nano Banana Pro can also generate text in multiple languages, allowing you to make multiple versions of your image for different markets. Google recommends checking the results as it can "make grammar mistakes of miss specific cultural nuances."

A translated image generated by Nano Banana Pro.
(Credit: Google)

All images can now be generated at either 2K or 4K resolution. You can select and refine any part of an image, such as changing a camera angle, color grading, effects, lighting, focus, and more.

Google's guidance suggests being as detailed as possible in your first prompt. It even recommends sharing specifics such as the aspect ratio to use, depth of field, and how you want lighting to appear.

Google has also expanded the tool to allow for 14 different images to be input at once, allowing you to bring more creations together in one complete AI-generated image. It says it can also maintain the resemblance of up to five subjects throughout the creation process. It means you can ask the tool to keep the people consistent even as you change other aspects, such as their poses.

An example pulled together from 14 different images.
(Credit: Google)

Nano Banana Pro is now available, with a usage limitation for non-subscribers. If you're on the free tier, you'll be dropped down to Gemini Flash 2.5 when you run out of credits. Those on a Google AI Plus, Pro, or Ultra subscription have higher quotas. The tool is also available within Google Search’s AI Mode in the US if you subscribe to either AI Pro or Ultra.

On the paid version of Google's NotebookLM, you’ll also get access to the AI-generation feature if you live in the US. Google will also bring these tools to its AI-filmmaking tool Flow, but an exact release date was not announced.

About Our Expert

James Peckham

James Peckham

Reporter

I’ve been a journalist for over a decade after getting my start in tech reporting back in 2013. I joined PCMag in 2025, where I cover the latest developments across the tech sphere, writing about the gadgets and services you use every day. Be sure to send me any tips you think PCMag would be interested in.

I’ve worked at TechRadar, Android Police, T3, and more, where I broke many tech stories you may have read, including the return of the Motorola Razr when it first became a foldable phone. Based near London, I’ve appeared on BBC News, Al Jazeera, and other TV networks, podcasts, and radio shows as an expert on the latest tech stories and trends.

Read full bio