Neo4j

Neo4j is an MCP server collection developed by Neo4j Labs that connects AI assistants to Neo4j graph databases and the Neo4j Aura cloud platform. Software engineers, data scientists, and knowledge engineers use this suite to run graph operations directly from natural language prompts inside MCP clients. The project provides multiple modular servers: mcp-neo4j-cypher inspects schemas and executes read or write Cypher queries; mcp-neo4j-memory maintains a persistent knowledge graph across chat sessions; mcp-neo4j-cloud-aura-api creates, resizes, and monitors Neo4j Aura instances; and mcp-neo4j-data-modeling designs, validates, and visualizes graph models with support for Arrows.app. These servers support STDIO, SSE, and HTTP transport layers, allowing local execution alongside client applications or remote containerized deployments on platforms like AWS ECS Fargate and Azure Container Apps. By exposing database introspection, query generation, and memory storage tools to LLMs, the servers enable automated graph traversal, dynamic data modeling, and context-aware reasoning over interconnected records.

Category: AI Memory & Context

Tags: cypher, graph database, knowledge-graph, neo4j

Visit Neo4j

How to install and configure Neo4j

1. Verify that your Neo4j instance is running and has the APOC plugin installed and enabled for schema inspection. 2. Review the repository at https://github.com/neo4j-contrib/mcp-neo4j to choose the specific server component you want to run, such as mcp-neo4j-cypher or mcp-neo4j-memory. 3. Configure your MCP client, such as Claude Desktop, VS Code, Cursor, or Windsurf, to execute the chosen server binary via STDIO or HTTP transport. 4. If running HTTP transport directly from the command line, run mcp-neo4j-cypher --transport http or configure environment variables such as export NEO4J_TRANSPORT=http along with NEO4J_MCP_SERVER_HOST, NEO4J_MCP_SERVER_PORT, and NEO4J_MCP_SERVER_PATH. 5. Restart your MCP client and verify that the tools appear in the assistant interface.

What you can do with Neo4j

• Executing Cypher Queries: Translate user requests into Cypher read and write statements to inspect database schemas, retrieve graph records, or update nodes and relationships automatically.

Key facts

• Open Source
• https://github.com/neo4j-contrib/mcp-neo4j
• AI Memory & Context, Databases & Data Stores
• cypher, graph database, knowledge-graph, neo4j

Part of MCP Servers

What is the Neo4j MCP server?

The Neo4j MCP server is an open source collection of Model Context Protocol servers created by Neo4j Labs. It allows AI clients like Claude Desktop to inspect graph schemas, execute Cypher queries, manage persistent knowledge graph memories, orchestrate Neo4j Aura cloud instances, and build interactive graph data models.

Which MCP clients work with Neo4j?

Neo4j works with any MCP-compliant client interface. Documented integrations include Claude Desktop, Cursor, VS Code, Windsurf, and the Gemini CLI. It connects locally through standard input and output or remotely via HTTP and Server-Sent Events transport layers.

What are the requirements for running mcp-neo4j-cypher?

To run mcp-neo4j-cypher, you need an accessible Neo4j instance with network connectivity. For schema inspection via the get_schema tool, the APOC plugin must be installed and enabled on the target database instance.

Is Neo4j MCP open source?

Yes, the Neo4j MCP servers are released as open source software under the MIT License on GitHub. Anyone can inspect the code, customize the tools, submit improvements, or deploy the containerized servers to cloud hosting providers.