Azure AI Search MCP Server

What is Azure AI Search MCP Server?

An MCP server that enables Claude Desktop to search your content using Azure AI services. It provides two implementations: the recommended Azure AI Agent Service (with document and web search) and direct Azure AI Search integration. Designed for developers and users who need AI-enhanced search capabilities within Claude.

How to use Azure AI Search MCP Server?

Install dependencies with uv and Python 3.10+, create a .env file with your Azure credentials, and run either azure_ai_agent_service_server.py or azure_search_server.py. Configure Claude Desktop by adding the server to its JSON config with the appropriate environment variables.

Key features of Azure AI Search MCP Server

• AI-Enhanced Search via Azure AI Agent Service
• Multiple Data Sources: private documents and public web
• Source Citations for web search results
• Flexible Implementation: Agent Service or direct Search
• Seamless Claude Desktop integration
• Customizable search behavior and tools

Use cases of Azure AI Search MCP Server

• Search indexed documents with AI-enhanced results from Claude
• Perform web searches with source citations through Bing
• Use keyword, vector, or hybrid search on Azure AI Search indices
• Combine private and public data sources in a single interface
• Prototype and extend search capabilities with custom tools

FAQ from Azure AI Search MCP Server

What is the difference between the two implementations?

The Azure AI Agent Service implementation (recommended) provides both document search and web search with AI-enhanced results. The direct Azure AI Search implementation offers keyword, vector, and hybrid search methods.

What are the requirements to run this server?

Python 3.10 or higher, the latest Claude Desktop, and Azure resources: an AI Search service with an index (for direct search) or an Azure AI Project with Search and Bing connections (for Agent Service).

How do I authenticate with Azure?

For the Agent Service implementation, run az login to authenticate. For the direct Search implementation, provide your API key in the environment variables.

Where does my data live?

All searches run against your own Azure AI Search index or Bing connection. No data is sent to third parties beyond your configured Azure resources.

Can I customize the server's behavior?

Yes. You can modify tool instructions, add new tools using the @mcp.tool() decorator, customize response formatting, and adjust web search parameters.