Google pulls out all the stops as Gemini "swallows" 250 million map data points, turning the directionally-challenged AI into a smart one overnight.

The world comes alive in the eyes of AI! Google's Grounding with Google Maps feature is launched. Gemini can access information on 250 million locations. Combined with search tools, it provides more accurate and timely answers, making it perfect for travel planning and local services.

Google pulls out all the stops!

Google has launched the "Grounding with Google Maps" feature, and Gemini is integrated with Google Maps.

In the past, language models understood text; now, they can understand "geospatial space".

Map + AI unlocks location intelligence.

Now, Gemini can directly access 250 million real - time geographical data points from Google Maps —

Enabling AI to have "geospatial understanding and reasoning abilities".

Google's move has created a literal "world model", making X users exclaim in admiration!

Anchored to Google Maps, AI understands the world

Since the 17th of this month, Google has launched the Google Maps tool in the Gemini API, allowing developers to anchor their applications to map data.

As a result, Gemini's reasoning capabilities can be connected to real - world data from over 250 million locations, giving rise to a new class of geospatial - aware AI products.

Just like Grounding with Google Search, Grounding with Google Maps can also provide rich and real - time updated data support for the model in location - related information queries.

Now, AI can know

What each place in the world is, where it is, what it has, when it opens, how to book tickets.... How to get there?

The function of this tool is:

When Gemini detects that your question involves geographical, location, route, business, or regional information, it automatically calls Maps data to answer.

This update is not only a technological enhancement but also an expansion of AI's capabilities.

This allows third - party developers to combine Gemini's reasoning capabilities with Maps' geographical data to generate more accurate and realistic responses —

This feature makes it out of reach for competitors such as OpenAI's ChatGPT, Anthropic's Claude, and many open - source models in the short term.

This integration is particularly suitable for local search, food delivery and distribution, real estate, and travel planning —

Any scenario that requires consideration of distance, real - time availability, or location personalization.

When the user's geographical location is known, developers can include latitude and longitude parameters in the request to improve the relevance and accuracy of the answer.

Google recommends enabling this tool only when the query clearly involves geographical context to save performance and cost.

According to the developer documentation, the pricing is $25 per 1000 requests — which is not cheap for high - frequency applications.

Has AI awakened geographical intelligence for a true "world model"?

Relying on Google Maps' vast global database, this new Gemini feature provides relevant and personalized content, thereby enhancing the user experience.

Four typical application scenarios:

Intelligent geographical Q&A: Answer questions about specific geographical locations comprehensively and accurately.

Route planning: Build conversational travel planning tools and local guides.

Personalized map recommendations: Recommend map points based on location and user preferences (e.g., restaurants or stores).

Intelligent LBS applications: Create location - based experiences for social, retail, or food delivery services.

In scenarios where geographical distance and current factual data need to be considered, such as finding "the best coffee shops near me" or getting a route, Google Maps' "grounding" (fact anchoring) function performs excellently.

Some developers have already used the new API to explore Berlin with Geminin and Google Maps:

Enterprises can use it for:

Business site selection: Help retailers find the best locations for new stores. It provides suggestions by analyzing demographic data, foot traffic, competitor situations, and local review sentiment.

Dynamic logistics and delivery optimization: Optimize routes in real - time, designed specifically for logistics companies. It can adjust according to real - time traffic, weather, and customer availability.

Real estate market analysis: Use geospatial data to evaluate property values, predict market trends, and identify investment opportunities.

Field staff management: Designed specifically for field service teams. It provides optimized routes and gives employees key customer and task information on - site.

Hyper - local marketing: A marketing tool that can send personalized, location - based offers to customers in real - time. For example, send a discount code when a customer passes by a store.

More features

To build more powerful and context - rich applications, in a single request, Gemini can now enable both Grounding with Google Maps and Grounding with Google Search:

Google Maps: Provides structured and verifiable factual data, such as addresses, business hours, and user ratings.

Google Search: Provides descriptive and timely context from the entire web, such as event schedules, news, and articles.

For example, when a user asks "Where can I find live music on Beale Street", Gemini can use both tools. The model can get the venue's business hours from maps and find the specific start time of the performance that night from search.

Google's internal evaluation shows that using both anchoring tools together significantly improves the quality of answers compared to using only one of them.

Currently, only specific Gemini models support fact anchoring based on Google Maps:

Gemini 2.5 Flash - Lite, Gemini 2.5 Pro, Gemini 2.5 Flash, and Gemini 2.0 Flash (excluding 2.0 Flash Lite).

To intuitively display the results of map - data anchoring, you can also use the returned context token to obtain an interactive widget.

It can also be rendered into the user interface (UI) of the application, providing users with a familiar experience and presenting relevant photos, user reviews, and other details.

Quick start

To start using it, enable the Grounding with Google Maps tool in your Gemini API request.

Here is an example using the Python SDK:

• from google import genai
• from google.genai import types
• client = genai.Client()
• prompt = "What are the best Italian restaurants within a 15 - minute walk from here?"
• response = client.models.generate_content(
• model='gemini-2.5-flash-lite',
• contents=prompt,
• config=types.GenerateContentConfig(
• tools=[types.Tool(google_maps=types.GoogleMaps())],
• # Optionally provide the relevant location context (this is in Los Angeles)
• tool_config=types.ToolConfig(retrieval_config=types.RetrievalConfig(
• lat_lng=types.LatLng(
• latitude=34.050481, longitude=-118.248526))),
• ),
• )
• print(response.text)
• if grounding := response.candidates[0].grounding_metadata:
• if grounding.grounding_chunks:
• print("Google Maps sources:")
• for chunk in grounding.grounding_chunks:
• print(f'- [{chunk.maps.title}]({chunk.maps.uri})')

Ask detailed questions about specific locations and get answers based on Google user reviews and other Google Maps data.

Get tailored recommendations based on user preferences and specific geographical regions:

Assist in trip planning and generate multi - day plans with route information and details of various locations, which is very suitable for travel applications.

In this example, by enabling the widget in the Google Maps tool, the user requested the googleMapsWidgetContextToken.

After enabling, the returned token can be used to render the contextual place widget using the