Google is making a very loud pitch to advertisers
AI image tools have always struggled with one thing: putting clear, readable text inside images.
Most models still produce typos or distorted fonts, which makes them unreliable for design work.
Google says its new model, Nano Banana Pro, finally fixes this.
It’s powered by Gemini 3, which the company describes as a big upgrade in reasoning and coding ability.
The launch pushed Alphabet’s share price to a record high, and it’s part of Google’s broader move to turn its AI tools into paid products.
Nano Banana Pro can generate sharper visuals with accurate text in multiple languages.
Free Gemini users will get access with limits, and paid users will have larger quotas.
The model is already being built into design tools such as Canva, Figma, Firefly and Photoshop.
Google says the model is better at planning layouts before it creates an image — choosing where the text goes, how it looks and how it fits with the rest of the design.
It can also turn things like recipes into neat diagrams, or create visuals using real-time information.
Quick recap:
Google’s new model focuses on fixing text quality in AI-generated images.
Nano Banana Pro runs on Gemini 3 and is being added to major design platforms.
Users can check whether an image was made with Google AI and customise outputs with detailed prompts.
AI learned typography, finally
Brands can upload up to 14 reference images, and the model will rework them into new scenes while keeping their original style.
Users can also specify things like camera angle, colour grading or aspect ratio.
Google is also adding a feature that lets people upload an image to the Gemini app and check if it was created by Google AI.
Support for audio and video will follow.
Every AI-generated file already includes invisible watermarking, and free or Pro images include a visible mark unless you’re on the Ultra plan.
Let’s come up with a serious name so people know how good this model is. Nano Banana Pro? Perfect. - MV