Neo4j MCP
by Neo4j Labs
Natural language interaction with Neo4j graph databases via multiple specialized servers
database Python Intermediate Self-hostable
⭐ 911 stars 📅 Updated: 1mo ago
Description
Neo4j MCP servers enabling natural language interaction with Neo4j graph databases. Includes four specialized servers: Cypher (translates natural language to Cypher queries with read/write execution), Memory (knowledge graph persistence for AI agent memory), Cloud Aura API (Neo4j Aura instance management), and Data Modeling (interactive graph modeling with Arrows.app integration). Supports STDIO, SSE, and HTTP transports for flexible deployment. Ideal for developers building AI applications that leverage graph-based knowledge representation and relationship-rich data.
✅ Best for
Teams using Neo4j who want AI-powered graph querying, knowledge graph memory, or cloud instance management
⏭️ Skip if
You use relational databases exclusively and don't need graph data modeling
💡 Use cases
- Querying Neo4j graph databases using natural language instead of raw Cypher
- Building persistent AI agent memory using knowledge graphs
- Managing Neo4j Aura cloud instances programmatically from AI assistants
- Interactive graph data modeling with visual Arrows.app integration
👍 Pros
- ✓ Four specialized servers covering queries, memory, cloud management, and modeling
- ✓ Supports STDIO, SSE, and HTTP transports for flexible integration
- ✓ Knowledge graph memory enables persistent, relationship-aware AI agent state
- ✓ Strong community backing from Neo4j Labs with active development
👎 Cons
- ✗ Requires an existing Neo4j database instance — no built-in provisioning for self-hosted
- ✗ Graph database concepts have a learning curve for relational database users
- ✗ Multiple server components may require separate configuration
🔧 Exposed tools (6 tools)
| Tool | Category | Description |
|---|---|---|
| manage_aura_instance | cloud | Manage Neo4j Aura cloud database instances |
| store_memory | memory | Store entities and relationships in the knowledge graph |
| retrieve_memory | memory | Retrieve stored knowledge from the graph memory |
| model_graph | modeling | Create and modify graph data models interactively |
| execute_cypher | query | Execute Cypher queries against a Neo4j graph database |
| inspect_schema | schema | Inspect the schema and structure of a Neo4j database |
⚡ Installation
Prerequisites:
- • python v3.10+
- • Neo4j credentials (URI, username, password)
- • API key required
Check Claude Code documentation to configure this MCP server.
💡 Tips & tricks
Start with the Cypher server for natural language queries against your existing Neo4j database. The Memory server is excellent for building AI agents with persistent knowledge — it stores entities and relationships as graph nodes and edges.
Quick info
- Author
- Neo4j Labs
- License
- MIT
- Runtime
- Python 3.10+
- Transport
- stdio
- Category
- database
- Difficulty
- Intermediate
- Self-hostable
- ✅
- Auth
- —
- Docker
- —
- Version
- latest
- Updated
- Feb 28, 2026
Client compatibility
- ✅ Claude Code
- ✅ Cursor
- ✅ VS Code Copilot
- ❓ Gemini CLI
- ❓ Windsurf
- ❓ Cline
- ❓ JetBrains AI
- ❓ Warp
Platforms
🍎 macOS 🐧 Linux 🪟 Windows