Multi-Model Database

Modern data doesn’t fit neatly into tables. IoT platforms generate time-series sensor data, applications log JSON payloads, search systems index text, and AI models rely on vector embeddings, all at once.

Traditional databases force you to separate this data into different systems. CrateDB eliminates that complexity by supporting multiple models natively inside one scalable, high-performance platform.

• Relational: classic SQL tables for structured business data
• Time-series: optimized for high-velocity event streams and monitoring metrics
• JSON / Object: flexible schema for semi-structured and evolving data
• Full-text: fast search and scoring with integrated text indexing
• Geospatial: spatial functions and indexing for location intelligence
• Vector: native support for embedding search and AI similarity queries

Everything is accessible through standard SQL, without middleware, ETL pipelines, or external indexes.

CrateDB’s shared-nothing, distributed SQL engine handles all data types uniformly. Each node stores data locally in shards (whether it’s a text field, a JSON object, or a vector) and the query engine parallelizes processing across the cluster.

• Unified schema management: mix and match data types within a single table or across joins.
• Automatic indexing: every field type is automatically indexed for high-speed queries.
• Hybrid queries: combine aggregations, filters, search, and vector matching in one SQL statement.
• Scalable ingestion: handle millions of records per second from diverse data sources.

With CrateDB, there’s no need to choose between SQL analytics, document queries, or search. You can do it all in one query.

CrateDB unifies these models into one scalable, real-time data platform. No connectors, no data silos.

Data Type	Example Use Case	CrateDB Capability
Time-series	IoT telemetry, sensor monitoring	High-throughput ingestion and real-time aggregations
JSON / Object	Application logs, event payloads	Flexible schema with automatic indexing
Text / Full-text	Product search, document tagging	Relevance-based ranking via MATCH
Geospatial	Fleet tracking, geofencing	Location queries with within(), distance(), and intersects()
Vector	AI embeddings, semantic search	Native KNN_MATCH() for similarity search
Relational	Customer and asset data	SQL joins, constraints, and aggregations