Corporate AI failure in 2025 it feels inevitable. From website banner ads to outdoor billboards, I am surrounded by images generated by companies using AI tools. Hell, even the bar down the street has content hour flyers posted with a distinctly hazy amber glow to some of the AI graphics.
On Thursday, Google launched the Nano Banana Pro, the company’s latest image-generating model. Many of the updates in this release are intended for enterprise adoption, starting with the inclusion of Nano Banana Pro in Google Slides business presentations to integrate the fresh model with Google Ads for advertisers around the world.
This ‘Pro’ edition is an iteration of the Nano Banana that launched earlier this year. Nano Banana became a viral sensation when users started posting personalized figurines and other creations that can be memed on social media.
Nano Banana Pro develops the AI tool with a number of fresh capabilities, such as generating images in 4K resolution. The Google Gemini app is free to try, and paid Google One subscribers get access to additional generations.
One specific improvement will be corporate catnip in this release: text rendering. From my initial tests generating text output, Nano Banana Pro improves the blurry letters and strange spelling errors common in many image models, including earlier versions of Google.
Google wants the images generated by this fresh model – text and all – to be more polished and business-ready. “Even if one letter is missing, it’s very obvious,” says Nicole Brichtova, product lead for images and video at Google DeepMind. “It’s a bit like having hands with six fingers: it’s the first thing you see.” He says one of the reasons why the Nano Banana Pro is able to generate text in a cleaner way is the switch to a more capable entry-level model, the Gemini 3 Pro.