Weaviate MCP Server

Overview

Weaviate's vector search capabilities can now be integrated through a Weaviate MCP server. The Model Context Protocol (MCP) provides a consistent method for providing context to AI models. The mcp-server-weaviate is designed to work with Weaviate vector search engine, enabling users to manage and query their data effectively.

Available Implementations

There are multiple MCP server implementations for Weaviate:

1. Official Weaviate MCP Server

Supports:

• Retrieving objects from Weaviate with hybrid search
• Inserting objects into Weaviate

Official repository: github.com/weaviate/mcp-server-weaviate

2. Community Implementation (FastMCP-based)

Provides seamless integration with Weaviate vector databases with 11 tools including:

• Configuration viewing
• Connection checking
• Collection listing
• Schema retrieval

Search Capabilities

Vector Similarity Search

Using embeddings for finding conceptually similar content

BM25 Keyword Search

Exact term matching for precise keyword queries

Hybrid Search

Combines semantic and keyword search using Reciprocal Rank Fusion (RRF) for optimal results

Key Features

• Powerful search capabilities across different search paradigms
• Schema management and object manipulation
• Integration with AI agents for context-aware retrieval
• Support for both semantic understanding and exact keyword matching

Use Cases

• Semantic search in knowledge bases
• Hybrid search for e-commerce catalogs
• Document retrieval for RAG applications
• Context-aware AI agent memory
• Multi-modal search across text and embeddings

Integration

Works with Claude Desktop, Cursor, and other MCP-compatible AI assistants for vector search operations.