Baidu Maps MCP Server Integration

Overview

Baidu Maps API is now fully compatible with the MCP Protocol, making it the first domestic map service provider to support MCP. The Baidu Maps MCP Server includes 10 API interfaces that comply with the MCP protocol standards, covering reverse geocoding, place search, route planning, and more. It is developed using the MCP Python SDK and MCP Typescript SDK, allowing any MCP-compatible intelligent assistant (such as Claude, Cursor, and Qianfan AppBuilder) to quickly integrate.

Tools

1. Geocoding (map_geocode)
2. Converts an address into corresponding coordinates.
3. Input: address

4. Output: location

5. Reverse Geocoding (map_reverse_geocode)

6. Converts coordinates into a semantic address.
7. Input: location

8. Output: formatted_address, uid, addressComponent

9. Place Search (map_search_places)

10. Searches for places and POIs in various scenarios, including city search and circular area search.
11. Input: query, location, radius, region

12. Output: POI list with name, location, address

13. Place Details (map_place_details)

14. Retrieves detailed information about a POI based on its uid.
15. Input: uid

16. Output: POI details including name, location, address, brand, price

17. Batch Route Calculation (map_distance_matrix)

18. Calculates the distance and travel time between multiple origin and destination points.
19. Input: origins, destinations, mode

20. Output: Travel time and distance for each route

21. Route Planning (map_directions)

22. Plans travel routes and estimates travel time based on origin and destination coordinates.
23. Input: origin, destination, model

24. Output: Route details including steps, distance, duration

25. Weather Query (map_weather)

26. Queries weather information based on administrative region codes or coordinates.
27. Input: district_id, location

28. Output: Weather information including temperature, weather, wind

29. IP Location (map_ip_location)

30. Retrieves the location of a request based on its IP address.
31. Input: ip

32. Output: Current city and city midpoint location

33. Real-Time Traffic (map_road_traffic)

34. Queries real-time traffic conditions based on road names or area shapes.
35. Input: model, road_name, city, bounds, vertexes, center, radius

36. Output: Traffic information including road_name, traffic_condition

37. POI Extraction (map_poi_extract)

    • Extracts POI information from text content (requires advanced permissions).
    • Input: text_content
    • Output: Related POI information including name, location

Getting Started

Obtaining an AK

Before using the Baidu Maps MCP Server, you need to create a server-side AK in the Baidu Maps Open Platform Console.

Python Integration

Refer to the Python Integration Documentation.

Typescript Integration

Node.js Installation

Install Node.js. Verify the installation by running:

node -v

Configuration

1. Open Claude for Desktop settings, navigate to Developer, and click Edit Config.
2. Add the following configuration to the file, replacing BAIDU_MAP_API_KEY with your AK:

{
    "mcpServers": {
        "baidu-map": {
            "command": "npx",
            "args": ["-y", "@baidumap/mcp-server-baidu-map"],
            "env": {
                "BAIDU_MAP_API_KEY": "xxx"
            }
        }
    }
}

1. Restart Claude to load the Baidu Maps MCP Server.

Example Usage

Ask Claude for travel planning, and it will use the MCP Server to provide detailed routes and travel suggestions.

Integration with Qianfan AppBuilder

Agent Configuration

1. Create and publish an application on the Qianfan Platform.
2. Set the Agent's thinking rounds to 6 and publish the app.

Example Code

import os
import asyncio
import appbuilder
from appbuilder.core.console.appbuilder_client.async_event_handler import (
    AsyncAppBuilderEventHandler,
)
from appbuilder.modelcontextprotocol.client import MCPClient

class MyEventHandler(AsyncAppBuilderEventHandler):
    def __init__(self, mcp_client):
        super().__init__()
        self.mcp_client = mcp_client

    async def interrupt(self, run_context, run_response):
        thought = run_context.current_thought
        print("\033[1;31m", "-> Agent Thought: ", thought, "\033[0m")
        tool_output = []
        for tool_call in run_context.current_tool_calls:
            tool_res = await self.mcp_client.call_tool(
                tool_call.function.name, tool_call.function.arguments
            )
            tool_output.append({"tool_call_id": tool_call.id, "output": tool_res})
        return tool_output

async def agent_run(client, mcp_client, query):
    tools = mcp_client.tools
    conversation_id = await client.create_conversation()
    with await client.run_with_handler(
        conversation_id=conversation_id,
        query=query,
        tools=tools,
        event_handler=MyEventHandler(mcp_client),
    ) as run:
        await run.until_done()

async def main():
    appbuilder.logger.setLoglevel("DEBUG")
    app_id = ""
    appbuilder_client = appbuilder.AsyncAppBuilderClient(app_id)
    mcp_client = MCPClient()
    await mcp_client.connect_to_server("./<YOUR_FILE_PATH>/map.py")
    await agent_run(appbuilder_client, mcp_client, 'Drive from Beijing to Shanghai')
    await appbuilder_client.http_client.session.close()

if __name__ == "__main__":
    loop = asyncio.get_event_loop()
    loop.run_until_complete(main())

Example Result

The Agent will use the MCP Server to retrieve navigation routes and provide travel suggestions.

Notes

• Administrative region codes use the Baidu adcode mapping table.
• Coordinates use the gcj02ll system. Refer to Baidu Coordinate System.

License

MIT License

Feedback

Report issues or submit feedback via Baidu Maps Open Platform.

Updates