Google Earth AI

Google Earth AI is an analytics tool. Google Earth AI combines Google's geospatial models for floods, wildfires, cyclones and imagery with Gemini reasoning, letting enterprises, nonprofits and cities query planetary data in natural language through Google Earth, Maps Platform and Google Cloud.

Google Earth AI is a collection of geospatial models and datasets that Google has built over years of modeling the planet, now combined with Gemini reasoning so that location questions can be asked in plain language. The models cover phenomena such as floods, wildfires, air quality and cyclones, alongside imagery models that work on satellite and aerial pictures. The aim is to shorten analysis that once demanded complex pipelines and long iteration into a matter of minutes. The offering is aimed at enterprises, nonprofits, city governments and researchers who need environmental monitoring, disaster response or public health planning. Example prompts on the homepage show the range: finding storm drains within 50 meters of public schools in a Brooklyn neighborhood, judging which communities in the Democratic Republic of Congo face the highest cholera risk, or locating mangrove forests that resemble those in the Mekong delta. Partners named by Google include Planet, Airbus, Deloitte, WPP, Boston Children's Hospital, McGill and Partners, and GiveDirectly, which use it for tasks like object detection by natural language, predicting property damage before storms and triggering cash assistance ahead of floods. In practice, Earth AI is not a single app. Its models power features in Google Earth, Google Maps Platform and Google Cloud. Inside Google Earth, an Ask Google Earth chat panel returns answers as data tables and map pins. Satellite providers access Remote Sensing Foundation Models through a trusted tester program, and developers can use geospatial analytics on Google Cloud's Gemini Enterprise Agent Platform, which grounds agents in Google Maps data covering more than 300 million places. Flood forecasting is cited as reaching more than two billion people. Among alternatives, it sits between traditional GIS suites and generic satellite analytics platforms. Its edge is the scale of Google's underlying data and models plus conversational access, while the breadth of entry points means buyers must work out which surface fits their use case.