Google goes bananas with its latest gen AI update – here are the imaging upgrades to expect

Google has shown off a new update to its AI image generation model Nano Banana 2, and it's looking fresh.

This is part of the Google Gemini AI sub-model that's specifically focused on the task of image creation based on requests.

The idea here is to bring the original Nano Banana model, released in August, more up to speed with the top-end Nano Banana Pro that landed in November. This latest Gemini 3.1 Flash Image, Nano Banana , is here to offer high-quality results at faster speeds.

What's new in Google Nano Banana 2?

One of the key offerings in this latest AI image generation model is speed. Since this now uses the high-speed smarts of Gemini Flash, image generation is faster than ever.

Crucially, this means quick iterations and edits can be done without lots of waiting. That should mean even if it isn't right first time, you can tweak and adjust to perfection quickly.

Accuracy is also important here as information is pulled from Gemini's real-world and real-time information base. That should mean more accurate renders with deeper understanding of what you're asking for.

Text is also more usable now, with precision addition, making this far more useful for marketing, greetings cards and mock-ups alike.

Creative control is improved with higher-quality photorealistic imagery available. Character resemblance can be maintained for up to 14 objects in a workflow. Better instruction understanding allows the model to follow commands, even when complex, with greater results.

You are also able to output to production ready quality, including up to 4K outputs, ideal for social posts or wide-screen backdrops. Thanks to more vibrant lighting and richer textures, details should be far more impressive also.

Google Nano Banana 2 is available to try and use now via the Gemini app, Google Search, AI Studio + API, Google Cloud, Flow and Google Ads.