- Google’s Nano Banana 2 AI image generator previews
- New model promises smarter, self-correcting image generation
- New model mimics a multi-step human workflow
Google is preparing for a new milestone in advanced AI image generation with Nano Banana 2. The follow-up to the original Nano Banana model will be part of the Gemini app, according to the preview that appeared unexpectedly and shared on X.
Nano Banana 2 seeks to bring greater control over the angle and viewpoint of images, with more precision in coloring. There will also be the option to correct the text in the image without disrupting the rest of the output.
So Nano Banana 2 will be more than just a resolution upgrade. Perhaps Nano Banana 2’s biggest change is in its thinking. The leaked previews suggest that the model now adopts a multi-step workflow where it will plan the image before creating it, analyze it for errors, correct them, and repeat the process until it is ready.
This type of built-in autocorrection is new to Google’s image tools. Behind the fruit’s original nickname, Google seems to treat AI image generation like a real design assistant, drawing drafts, spotting flaws, and only handing you the final product after making sure it’s not terrible.
Nano Banana 2 is now in preview đź‘€currently on Media IO pic.twitter.com/VNmQM3PAFPNovember 8, 2025
Images shared by those with access to the preview show sharper lines, sharper angles, and fewer of the telltale errors of AI images.
“Nano Banana Pro” has started showing up in GitHub commits and code references, suggesting that Google is already preparing a more premium version of the model for high-end or high-resolution tasks.
Attractive AI Images
In the wild, Nano Banana 2, known internally as GEMPIX 2, has also started to appear beyond the main Gemini app. Testers found traces of the model in experimental tools such as Whisk Labs, part of Google’s ongoing efforts to integrate AI creativity into all of its tools. If the past is any indication, this multi-surface rollout will follow the same playbook as the original Banana launch, with users suddenly discovering that their photos look better.
The fact that Nano Banana 2 goes through a self-correction loop before delivering the final image marks a philosophical shift. It’s no longer just reactive. Google teaches its model to notice its own errors and proactively correct them. It seems small, but it reflects the human creative process more than older AI tools.
Better understanding of the scene, better angle control and better text clarity should mean you’re more likely to get the image you actually want. Considering how Nano Banana went viral for making it action figure versions of people that looked like real products, it will be worth watching to see how well Nano Banana 2 pulls off similar images set in the real world. Judging by some of the images shared and how consistently they maintain that character likeness, you’ll be seeing some truly realistic and unreal depictions of people and places very soon.
Follow TechRadar on Google News And add us as your favorite source to get our news, reviews and expert opinions in your feeds. Make sure to click the Follow button!
And of course you can too follow TechRadar on TikTok for news, reviews, unboxings in video form and receive regular updates from us on WhatsApp Also.
The best business laptops for every budget
