Neo4j has released Relate, an AI-native open-source in-memory and in-process embedded database built in Rust for AI agents and agentic systems. It installs quickly, starts in milliseconds, and can run in Jupyter notebooks, browsers, containers, mobile, and edge devices. Relate brings multi-hop reasoning and GraphRAG to agents through full Cypher support, reducing hallucinations. It auto-detects agent harnesses, returns structured errors with retryable flags, and enforces row caps with truncation markers to avoid context-window saturation. Connectivity includes a built-in Model Context Protocol server, CLI, Bolt, and HTTP Query API endpoints. The engine delivers sub-millisecond in-memory reads and can combine vector similarity search with graph traversals in a single Cypher statement. Optional durable storage uses a write-ahead log. Relate is released under the Apache 2.0 license.
Neo4j has released Relate, an AI-native open-source in-memory and in-process embedded database built in Rust for AI agents and agentic systems. It installs quickly, starts in milliseconds, and can run in Jupyter notebooks, browsers, containers, mobile, and edge devices. Relate brings multi-hop reasoning and GraphRAG to agents through full Cypher support, reducing hallucinations. It auto-detects agent harnesses, returns structured errors with retryable flags, and enforces row caps with truncation markers to avoid context-window saturation. Connectivity includes a built-in Model Context Protocol server, CLI, Bolt, and HTTP Query API endpoints. The engine delivers sub-millisecond in-memory reads and can combine vector similarity search with graph traversals in a single Cypher statement. Optional durable storage uses a write-ahead log. Relate is released under the Apache 2.0 license.