ArangoDB Multi-Model Database

ArangoDB is an open-source native multi-model database combining graph, document, key-value, and full-text search in a single engine. Founded in 2012 by Claudius Weinberger and Frank Celler, ArangoDB has over 13,000 stars as of 2026. Unlike polyglot persistence requiring separate databases for different models, ArangoDB provides all models natively, enabling complex queries spanning multiple models in a single transaction. Key features include: document store (JSON-like documents with dynamic schemas), graph database (vertices and edges with native graph traversal, shortest path, and graph algorithms via AQL GRAPH functions), key-value store (simple lookups optimized for speed), AQL query language (declarative SQL-like language supporting joins, traversals, aggregations, and full-text search across all models), ACID transactions (multi-document, multi-collection with snapshot isolation), distributed cluster (sharding across DBServers with automatic failover and smart rebalancing), full-text search (ArangoSearch with inverted index, tokenization, stemming, and BM25 scoring), geo-indexing (point and polygon queries), RocksDB storage engine (LSM-tree for high write throughput), SmartGraphs (optimizing graph traversals in sharded clusters), Pregel graph processing (distributed algorithms like PageRank and Connected Components), Foxx microservices (REST APIs inside the database using JavaScript), and Kubernetes Operator deployment.