SurrealDB

SurrealDB is a multi-model database designed as a unified context layer for AI agents and modern applications. It consolidates documents, graphs, vectors, and time-series data into a single system, eliminating the need to orchestrate multiple specialized databases. Its goal is to simplify data access and state management from storage to in-memory usage with a single transactional model and minimal infrastructure overhead.

SurrealDB supports flexible data modeling with SQL-like querying, graph relationships, and document-style records in the same schema. It includes native vector indexing and search, enabling similarity queries for embeddings used in retrieval-augmented generation and other AI workflows. The database offers real-time change feeds and live queries, making it suitable for applications that require continuous synchronization of state across agents, services, and clients. Built-in authentication, access control, and row-level permissions reduce reliance on external middleware and custom APIs, while support for both server and embedded deployments allows it to run in the cloud, on the edge, or locally.