Memgraph documentation

Memgraph is a high-performance, in-memory graph database that powers real-time AI context. It serves as the graph engine for GraphRAG pipelines, AI memory systems and agentic workflows. The same architecture drives real-time graph analytics, such as fraud detection, network analysis and infrastructure monitoring.

Get started

Start Memgraph and run your first query in a few minutes, by hand or with your coding agent. To let the agent write Memgraph code from the current docs, give it the Agent Skills and the docs MCP server.

Build AI applications

Use Memgraph as the graph behind retrieval, memory and planning in AI systems.

Analyze graphs

Run PageRank, community detection, path finding and centrality on your graph, or write your own procedures.

Query data where it lives

Run graph queries over data that stays in PostgreSQL, ClickHouse, Snowflake or a data lake, without moving it into Memgraph.

Bring your data into Memgraph

Import data from files, migrate from Neo4j or a relational database, or stream it in from Kafka and Pulsar.

Connect your application

Query Memgraph from Python, JavaScript, Java, Go, Rust, C# and other languages over the Bolt protocol.

Run Memgraph in production

Deploy Memgraph on Docker, Linux, Kubernetes or a cloud provider, then manage users, backups, monitoring and high availability.

Reference

Concepts, data modeling, Cypher, the visual interface and every release.

If you get stuck, ask on Discord.