New tool lets users reimagine real-world satellite and 3D map views with AI-generated prompts, now live globally on the web
Google has integrated its Nano Banana 2 AI image generation model into Google Earth, letting users transform real satellite, aerial and 3D map views into custom AI-generated visuals. The feature is available globally on Google Earth for the web, allowing users to select any location, click “create image,” and enter a prompt to reimagine the landscape.
Reviving history and building infographics
One use case lets people visualise historical sites as they once looked, such as turning present-day ruins into a bustling ancient streetscape. Another application generates instant educational infographics by combining Gemini’s research capabilities with Nano Banana’s image rendering, letting users learn key facts about landmarks directly on the map.
Real estate pitches and home design previews
The tool also has practical applications for professionals and homeowners alike. Architects, developers and realtors can turn vacant plots into polished client pitches by generating renderings of proposed developments on real terrain. Homeowners and designers can similarly preview architectural concepts, such as a new structure built on their own property, rendered photorealistically within the actual landscape.
A sci-fi spin on familiar landmarks
For more experimental use, the tool allows users to reimagine well-known locations in entirely different aesthetics such as transforming a real corporate campus into a futuristic setting with glowing walkways and glass domes, while keeping the underlying real-world geography intact.
Grounded in real geography
According to Google, anchoring the AI generation to actual geographic data opens up creative, educational and professional applications that go beyond typical image generation tools, since the outputs are tied to real locations rather than generic scenes.
