We introduced two new methods in experimental, enabling Maps Grounding Lite to resolve entity names in third-party data sources to Google Maps Place IDs, allowing for the rendering of rich, factual data widgets.

Learn more in the API references.

We are introducing two experimental developer tools to the Weather API:

• Minute Forecasting Endpoint: Provides highly localized, minute-by-minute precipitation nowcasts with a look-ahead window of up to 6 hours. The data includes precipitation probability, intensity, type (rain, snow, hail), and accumulation.

• Weather Maps: High-fidelity, real-time visual raster layers that overlay directly onto Google Maps. These tiles display current precipitation conditions styled to match the industry standards.

We are excited to announce that JSON Import and Export is now Generally Available for Cloud-Based Maps Styling across Maps SDKs. This update enables designers, developers and GenAI tools to effectively treat map styles as code.

What's New?

The update brings a new modern, human readable format, version control support and the ability to easily share map styles across teams.  There’s legacy style auto-conversion, AI-ready workflows and expanded capabilities for POI pin layering.  Learn More

The Transit Data in Places API (New) is now generally available to developers, providing transit data for transit POIs. This answers a long-standing feature request from the external developer community. By requesting the transitStation field via a strict FieldMask, developers can now query public transit hubs directly within Place Details (New), Text Search (New), and Nearby Search (New).

Key Capabilities:

• Returns station-specific fields such as Station Name and Stop Id.

• Provides a list of transit lines serving the station, including Line Name, Color, Icon, and Type (e.g., Subway, Bus).

We are announcing the General Availability of the Detected Arrays feature within the Solar API. This enhancement provides a binary detection status (yes/no) indicating the presence of existing rooftop photovoltaic (PV) installations, alongside the timestamp of the source imagery utilized for the detection. This data is now integrated into the buildingInsights response payload, enabling developers to optimize their solar design processes.

If you couldn't join us in person or tune in live, don't worry - we’ve put together a complete recap of everything that was announced. We're diving deep into how we are bridging the gap between large language models and the physical world. We released our I/O Developer Session Building agents with real-world reasoning. In this session, Ken Nevarez and Caio Moreira from our Developer Experience team help you understand how Maps Grounding Lite uses the Model Context Protocol (MCP) to make it easy to connect your AI applications with reliable real-world geospatial data directly from Google Maps.

Read our full Google I/O recap and watch the session below:

A huge thank you to our Innovators for joining our Google Maps Platform Fireside Chat on May 15th! Our speakers, Developer Experience engineers Ryan Baumann and Teresa Qin, shared an insightful and engaging discussion on grounding your agentic solutions with accurate real world data from Google Maps Platform, including a demo of a real-world application of Google Maps Grounding Lite.

The recording is available on YouTube. Our February Fireside Chat, Build maps with AI is also available. Be sure to stay on the lookout for more Fireside Chats coming your way soon and if you haven’t already, join our Discord Community!