Nano Banana 2 Is Coming to Google Earth
You can now use Nano Banana 2 on Google Earth on the web to imagine historical scenes and futuristic developments.
Tekef
David Nield Freelance Writer
Experience
David Nield is a technology journalist from Manchester in the U.K. who has been writing about gadgets and apps for more than 20 years.
He has a bachelor's degree in English Literature from Durham University, where he also spent a term as editor of the award-winning student newspaper Palatinate. His journalism career started in print media, where he contributed to and edited several technology magazines and bookazines sold in the U.K. and internationally.
More recently, he has worked as a freelancer for some of the biggest technology publications on the web, covering everything from on-the-ground reporting about product launches, to detailed explainers and how-to guides on apps, gadgets, and platforms. His expertise covers broad areas of consumer tech, including smartphones, laptops, wearables, and AI.
July 30, 2026
Add as a preferred source on Google
Credit: Google
Key Takeaways
Google Earth is getting a Nano Banana 2 upgrade. With it, you can create futuristic vistas or dive back into the historic past of a place. The update is rolling out now for all Google Earth users on the web.Table of Contents
Known by the codename Nano Banana 2, the most recent AI image generation updates added to Google Gemini have been a significant step forward in terms of quality and accuracy, and now similar capabilities are being plugged into the Google Earth app on the web. It means you're able to use the aerial, satellite, and 3D imagery packed into Google Earth as the basis for prompts for AI images—maybe peering into the distant past with a historical reconstruction or looking to the future with a sci-fi reimagining.
We've already seen just how precise and realistic the Nano Banana-powered features in Google Gemini are now, and Google has five suggestions for how you might deploy these capabilities in Google Earth. First, you can learn while exploring, so maybe getting the AI to build an infographic on the Statue of Liberty with key historical facts about it.
One idea for the new feature: Infographics (the Statue of Liberty in this case). Credit: Google
A text prompt such as "reimagine this empty lot in Tokyo as a vibrant shopping and retail district with open spaces" would transform a rather bare part of the Google Earth 3D map into something quite different. You can also visualize personal projects such as a dream lakeside home or an extension for your current property. Google Earth would combine the existing landscape with whatever you prompted, and help you come up with ideas as part of the planning stage.
You can also look forwards and backwards. "Sometimes it’s fun to let your imagination run wild," says Google, suggesting that you might want to get the AI to imagine how a place might look in a hundred years. (It might be worth nudging your prompt towards a utopian rather than dystopian vision of the future.)
Finally, Google suggests you dive back into the past. The company showcased a demo with the prompt "render a hyper-realistic view of what these Pompeii ruins looked like in 78 AD," which could be used by teachers, for example. That would instantly transform today's ruins "into a bustling, colorful street scene from the Roman Empire," in Google's words.
What do you think so far?
These new features come with the usual AI image caveats, however. While Nano Banana 2 can pull real-world information and images from the web, and is able to render graphics with more consistency and accuracy than previous models, there's still going to be the usual weaknesses involved with AI: Your lakeside home might not be entirely structurally sound, and the Pompeii reconstruction may not be entirely authentic—so bear that in mind when your images are produced.
Pompeii in all its original glory, as depicted by AI. Credit: Google
When I asked Google about this, a representative emphasized the real-world knowledge base that Nano Banana 2 draws on, which includes all the text and graphics it can find online, as well as the vast number of images that have been plugged into the AI during its training. With Google Earth, it will also trace over the structure ruins that are still in place when it comes to Pompeii, for example.
For a bit more clarity, you can try quizzing the Nano Banana 2 model inside the main Gemini app: I got it to render an imagined scene from Pompeii just prior to it being smothered by Vesuvius, then asked about the image accuracy. The AI said it had combined a bunch of real, historical information points to render the scene, including details on Pompeii's raised sidewalks, and the clothing styles of the day. Google hasn't said anything specific about usage limits, but the Nano Banana 2 upgrade is rolling out now globally, for all users of Google Earth on the web.
