Get Started with ArcadeDB
ArcadeDB is a multi-model database management system: graph, document, key/value, time-series, vector, and geospatial data on the same storage, in the same transaction.
Query with SQL, Cypher, Gremlin, GraphQL, MongoDB QL, or Redis. Embeddable in any JVM, with built-in HA replication. Apache 2.0 licensed.
Multi-Model
Graph, Document, Key/Value, Search, Time Series, Vector, and Geospatial in one engine.
Multi-Language
SQL, Cypher, Gremlin, GraphQL, MongoDB QL, and Redis commands.
Multi-Protocol
HTTP/JSON, PostgreSQL, MongoDB, and Redis wire protocols.
Performance
Built for speed with minimal GC pressure and efficient storage.
ACID Transactions
Full transaction support with isolation and durability guarantees.
High Availability
Leader-follower replication with automatic failover.
Choose your starting point
Run ArcadeDB
Docker, native binaries, or Kubernetes — pick the deployment that fits your environment.
ArchitectureMulti-Model
Combine graph, document, key/value, vector, and time-series in a single transaction.
Java embeddedEmbedded Server
Run ArcadeDB inside your JVM application — no separate server process required.
Java · 10 minJava Quickstart
Hands-on tutorial: schema, transactions, graph queries, and the async API.
PythonPython Quickstart
Connect from Python over the PostgreSQL wire protocol or the HTTP API.
JavaScriptJavaScript Quickstart
Run queries from Node.js or the browser via HTTP and JSON.
Vector searchVector Search Tutorial
Build an HNSW index and run cosine-similarity queries for semantic search.
Time seriesTime Series Tutorial
Ingest high-cardinality metrics and query them with PromQL or SQL.
Connect from your application
ArcadeDB exposes several wire protocols and drivers, so you can keep your existing client libraries:
-
HTTP / JSON REST API — language-agnostic, no driver required
-
PostgreSQL wire protocol — existing Postgres clients connect unchanged
-
Neo4j Bolt protocol — reuse Neo4j’s official drivers
-
MongoDB QL — MongoDB drivers
-
Redis commands — Redis clients
-
MCP server — for AI agents and LLM tooling
-
Native Java API — embedded mode
-
JDBC driver — for BI and traditional SQL tools
Pick a query language
ArcadeDB supports six query languages over the same data — use whichever your team already knows:
-
SQL — extended with graph, full-text, and vector functions
-
Cypher (openCypher) — drop-in replacement for Neo4j Cypher
Next steps
-
Records, Documents, Vertices & Edges — how ArcadeDB models data
-
Schema — types, properties, constraints
-
Transactions — ACID guarantees
-
Self-managed deployment — production install
-
ArcadeDB Academy — free, self-paced courses with certification