Google Maps Embed API Tool

[VaibhavLakshmiS]

Checked

• I searched existing ideas and did not find a similar one
• I added a very descriptive title
• I've clearly described the feature request and motivation for it

Feature request

We aim to integrate the Google Maps Embed API into LangChain, enabling it to:

• Generating Google Maps embed URLs for a specified location, enabling users to visualize places through interactive maps.
• Customizing the map display with parameters such as zoom level and map types (e.g., roadmap, satellite, hybrid, terrain).

Motivation

In a tech-driven world, users increasingly depend on interactive geographical data to enhance their digital and real-world experiences. Integrating the Google Maps Embed API into LangChain would fill a critical gap by offering dynamic, real-time maps directly within the conversational interface, improving decision-making for location-based queries, such as routes and points of interest. This feature would complement the existing Google Places API, providing seamless map displays that support better planning and exploration through visually intuitive, familiar content for users.

Proposal (If applicable)

Our proposal is to develop a toolkit that leverages the Google Maps Embed API to generate map URLs based on user queries. This will involve building a utility that dynamically constructs embed URLs, along with a toolkit to integrate map display functionality within LangChain agents.

We anticipate making the following additions to the codebase:

• libs/community/langchain_community/tools/google_maps_embed for the toolkit.
• libs/community/langchain_community/utilities/google_maps_embed.py for the API wrapper.
• libs/langchain/langchain/tools/google_maps_embed for the toolkit.
• libs/langchain/langchain/utilities/google_maps_embed.py for the API wrapper.

We also anticipate modifying the following files:

• libs/community/langchain_community/tools/__init__.py to register the tool.
• libs/community/langchain_community/utilities/__init__.py to register the tool.

We are a team of three students from the University of Toronto, aiming to develop this feature and submit a pull request by mid-November.This integration seeks to expand LangChain’s capability to handle a wider range of location-based queries and enhance user experience through interactive mapping.