Grounding with Google Maps is now generally available in the Gemini API, enabling developers to deliver more accurate, context-rich, and visually engaging experiences powered by real-world location data. This feature allows models to automatically detect geographical context in user queries and ground their responses using live, reliable data from Google Maps. With this capability, developers can enhance applications across travel, real estate, retail, and logistics by providing detailed, location-aware answers. The model leverages structured data from Google Maps—including business details, addresses, operating hours, user reviews, photos, and nearby points of interest—to generate responses that are both factual and insightful. One key enhancement is the ability to use the returned context token to retrieve an interactive widget. This widget can be embedded directly into your application’s UI, offering users a familiar, rich experience with visual elements like photos, reviews, and key details—without leaving the app. Use cases include: Detailed itinerary planning: Go beyond simple lists. A travel app can generate a full day’s plan with optimized routing, travel times, and up-to-date business hours. For example, a user asking, “Plan a day in San Francisco for me. I want to see the Golden Gate Bridge, visit a museum, and have dinner,” receives a complete, practical itinerary with location-specific insights. Hyper-local personalized recommendations: Tailor suggestions based on user preferences and precise locations. A real estate platform could identify neighborhoods with nearby schools, parks, and playgrounds—ideal for families—by analyzing local amenities. Local place-based answers: Answer specific questions about a location using verified data. A query like, “Does the cafe on the corner of 1st and Main have outdoor seating?” can now be answered accurately using real-time Maps data. For even greater depth, developers can combine Grounding with Google Maps and Grounding with Google Search in a single API request. This dual approach delivers the best of both worlds: Google Maps provides structured, factual data such as addresses, hours, ratings, and venue details. Google Search offers dynamic, descriptive context from the web—such as event schedules, news updates, and articles. For example, when a user asks about “live music on Beale Street,” the model can use Maps data to confirm venue hours and Search data to identify the exact start times of evening performances. Internal evaluations show that combining both tools significantly improves response accuracy and relevance compared to using either one alone. Grounding with Google Maps is now available for use with the latest Gemini models, giving developers flexibility to balance performance, cost, and quality. To get started, explore the official documentation, remix the demo app in Google AI Studio, or dive into practical examples in the Gemini API Cookbook. Begin building smarter, more grounded applications today.