FalkorDB

active

by FalkorDB Ltd. · since 2023

Low-latency graph database using sparse matrices, the maintained successor to RedisGraph.

graphvectordatabase enginesource-availablegrowing

Overview

FalkorDB is a property-graph database that represents adjacency as sparse matrices and evaluates Cypher queries with linear algebra (GraphBLAS). It started in 2023 as a fork of RedisGraph after Redis announced that product's end of life, runs as a Redis-compatible server and speaks openCypher, and its core has since been rewritten in Rust for the 5.x/6.x line. It targets GraphRAG and agent memory, with vector and full-text indexes, multi-tenant graphs and a managed cloud. Licensing is SSPL, so it is source-available rather than open source.

Strengths & trade-offs

Strengths

  • +Sparse-matrix (GraphBLAS) execution gives very low traversal latency
  • +Drop-in path off RedisGraph with the same Redis-protocol command model
  • +Thousands of isolated graphs per instance suit multi-tenant SaaS and per-user agent memory
  • +Vector and full-text indexes alongside Cypher for GraphRAG

Trade-offs

  • −SSPL license restricts offering it as a managed service and is not OSI-approved
  • −A single graph must fit in one node's memory
  • −Cypher coverage is narrower than Neo4j's
  • −Young company and fast-moving codebase after the Rust rewrite

Great for

  • +GraphRAG knowledge graphs for LLM apps
  • +Agent and user memory graphs
  • +Migrating off RedisGraph
  • +Real-time recommendation and fraud lookups

Think twice for

  • −Single graphs larger than RAM
  • −RDF/SPARQL workloads
  • −Teams needing permissive licensing

Capabilities

Native or official first-party support · 7 of 25 tracked

✓Secondary indexes
✓Full-text search
✓Vector search
✓Geospatial
✓In-memory
✓Durable persistence
✓Role-based access
–Transactions
–Joins
–JSON
–Change data capture
–Triggers
–Stored procedures
–Materialized views
–TTL / expiry
–Pub/Sub
–Columnar storage
–Compression
–Encryption at rest
–Time travel
–Branching
–Online schema changes
–Streaming ingest
–Offline sync
–Runs in WebAssembly

Scale & consistency

Comfortable data size
Medium · ≤ 10 TB
Horizontal scaling
None
Multi-region
No
Transactions
Limited ACID
Consistency
Strong
Schema
Schemaless
Architecture
In memory, Primary replica, Client server
Query languages
Cypher

Each graph lives in memory on one shard; many independent graphs (multi-tenancy) spread across a cluster.

Deployment

Self-hosted Vendor-managed cloud Serverless Embedded Edge Kubernetes operator

available on · AWS · Google Cloud

Ecosystem

compatible withRedis
apis & protocolsRedis protocol (RESP), Bolt (experimental)
client languagesPython, JavaScript, Java, Go, Rust, C#

Alternatives to FalkorDB

Compare all→
RedisGraphGraph9.4

Retired Redis module that ran Cypher over sparse adjacency matrices; succeeded by FalkorDB.

falkordb vs redisgraph →
MemgraphGraph24.4

In-memory, Cypher-compatible graph database in C++ for streaming and real-time graph workloads.

falkordb vs memgraph →
Neo4jGraph61.2

The default property-graph database, with Cypher, a big ecosystem and a managed cloud.

falkordb vs neo4j →
KùzuGraph11.2

Embedded columnar property-graph database with Cypher; archived in 2025 after Apple acquired the team.

falkordb vs kuzu →
LadybugDBGraph20.7

Community-maintained continuation of Kùzu: embedded columnar graph database with Cypher.

falkordb vs ladybugdb →
HelixDBGraph13.2

Rust graph-vector database on object storage, queried through typed SDKs for AI apps.

falkordb vs helixdb →

FalkorDB FAQ

What is FalkorDB?

FalkorDB is a graph database by FalkorDB Ltd., first released in 2023. FalkorDB is a property-graph database that represents adjacency as sparse matrices and evaluates Cypher queries with linear algebra (GraphBLAS). It started in 2023 as a fork of RedisGraph after Redis announced that product's end of life, runs as a Redis-compatible server and speaks openCypher, and its core has since been rewritten in Rust for the 5.x/6.x line. It targets GraphRAG and agent memory, with vector and full-text indexes, multi-tenant graphs and a managed cloud. Licensing is SSPL, so it is source-available rather than open source.

Is FalkorDB open source?

Not in the OSI sense. FalkorDB is source-available: you can read the code, but the license restricts some uses, typically offering it as a competing managed service. License: Server Side Public License v1.

Does FalkorDB support ACID transactions?

Partly. FalkorDB offers limited ACID guarantees (for example, single-record or single-partition transactions), with strong consistency.

Does FalkorDB scale horizontally?

FalkorDB does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at medium scale (≤ 10 TB).

How can I run FalkorDB?

You can self-host it and use the vendor's managed cloud. Managed offerings include FalkorDB Cloud. It is available on AWS and Google Cloud.

What is FalkorDB best for?

FalkorDB is a strong fit for GraphRAG knowledge graphs for LLM apps; agent and user memory graphs; and migrating off RedisGraph. Think twice for single graphs larger than RAM.

What are the best alternatives to FalkorDB?

Common alternatives to FalkorDB include RedisGraph, Memgraph, Neo4j, Kùzu and LadybugDB. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.

How popular is FalkorDB?

FalkorDB ranks #97 of 844 databases on the WhichDB Score (28.5/100), and #4 among graph databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.