Google is expanding its Gemini API toolkit again with the introduction of Grounding with Google Maps. The new feature lets developers connect Gemini’s reasoning capabilities with Google’s huge collection of location data, which includes more than 250 million places around the world.
This addition gives AI applications geographic awareness. Instead of relying only on text-based logic, apps can now respond with grounded, real-world information pulled straight from Google Maps. That means travel assistants that can generate full itineraries, logistics bots that account for drive times, and real estate apps that pinpoint kid-friendly neighborhoods, all powered by real-time map data.
Developers can enable the feature using a simple API call in Python, as Google demonstrated in its documentation. Once turned on, Gemini can answer location-based queries like “What are the best Italian restaurants within a 15-minute walk from here?” and return grounded answers complete with links to Maps sources.
Google says the integration also enables interactive widgets, so developers can embed real Google Maps experiences directly into their app interfaces. These widgets can display reviews, business hours, ratings, and photos, giving users a familiar Maps-style experience without ever leaving the app.
For even better context, developers can combine Grounding with Google Maps and Grounding with Google Search in the same API call. This setup gives Gemini access to both structured factual data such as addresses and hours, and dynamic web content like event listings or news. Google claims its internal testing shows this combination greatly improves response accuracy and relevance.
Grounding with Google Maps is available now through the Gemini API and supported by Google’s latest models, including Gemini 2.5 Flash Lite. Developers can explore examples in Google AI Studio or remix the official demo app to start building location-aware AI experiences today.
