SurrealDB

Scalable, distributed document-graph database that unifies documents, graphs, vectors, and time-series in a single ACID transaction. Combines flexible data modeling with built-in real-time subscriptions, authentication, and API generation, so developers can build full backends with one system instead of stitching document stores, graph databases, and vector indexes together.

SurrealDB differentiates by offering multi-model querying in one engine: graph traversals, vector similarity search, and temporal filtering in a single SurrealQL statement. Enterprises like PolyAI and Permit.io use it to consolidate backends that previously relied on Neo4j, PostgreSQL, and separate vector stores. Compared to alternatives, it provides live queries and row-level permissions without external middleware, plus SDKs for Rust, JavaScript, Node.js, WebAssembly, Go, Python, Java, .NET, and PHP.

Key capabilities:

• Documents, graphs, vectors, and time-series in one SurrealQL query
• ACID transactions across all data models
• Built-in auth, row-level security, and API endpoints
• Live queries that push changes to subscribers in real time
• Embeddable in Rust, JavaScript, Python, .NET, and Go
• Runs in-memory, on RocksDB or SurrealKV, or as a multi-node cluster

Typical applications include context-aware AI agents with unified memory and knowledge graphs, real-time collaboration and dashboards using live queries, and apps with complex relationships that benefit from graph-native traversal. Developers migrating from PostgreSQL, MongoDB, or Neo4j can often replace multiple backend components with a single SurrealDB instance.