FalkorDB
Ultra-fast Graph Database for Generative AI and GraphRAG
FalkorDB is a Redis-based, low-latency graph database built on a sparse adjacency matrix representation. It supports the OpenCypher query language with proprietary extensions and provides native GraphRAG capabilities for AI applications.
Built for teams working with complex, interconnected data, FalkorDB delivers fast traversal and flexible querying in production environments.
Core Capabilities
Property Graph Model
Native nodes, relationships, and properties built on the openCypher model.
OpenCypher + Extensions
Full OpenCypher support with performance-focused enhancements.
High-Performance Storage
Sparse adjacency matrix representation for fast traversal at scale.
RESP + Bolt Protocols
Connect using RESP or Bolt for broad driver support.
GraphRAG Ready
Native integration with the GraphRAG SDK for LLM applications.
Choose Your Path
Graph Database
Build fast, production-ready graph apps with OpenCypher and robust indexing.
- Best for: Real-time traversal
- Start with: data model and first queries
GraphRAG & AI
Power GenAI workflows with graph-aware retrieval and reasoning.
- Best for: RAG pipelines, knowledge graphs
- Start with: GraphRAG SDK
Not sure? You can do both—start with the database and add GraphRAG later.
Quick Start
Launch a local instance:
docker run -p 6379:6379 -p 3000:3000 -it --rm falkordb/falkordb:latest
Alternative deployment options:
- FalkorDB Cloud - Managed service with free tier
- Kubernetes deployment - Production orchestration
- Railway - One-click deployment
Use Cases
Common scenarios include:
| Use case | Job to be done | Examples |
|---|---|---|
| Graph Analysis | Identify patterns and influence in connected data | Social networks, supply chain modeling, network topology |
| Knowledge Graphs | Organize entities and relationships for discovery | Entity relationship modeling, semantic search, data discovery |
| Recommendation Systems | Personalize results using relationship signals | Collaborative filtering based on relationship patterns |
| Fraud Detection | Detect suspicious patterns across networks | Pattern matching across transaction networks |
| GraphRAG | Ground LLM answers with structured context | Retrieval augmented generation with structured knowledge graphs |
| Master Data Management | Maintain a unified view of critical entities | Complex entity relationships and hierarchies |
Documentation Paths
Graph Database
For developers using FalkorDB as a property graph database:
- Getting Started Guide - Installation and first queries
- OpenCypher Reference - Query language syntax and features
- Client Libraries - Python, JavaScript, Java, Rust, Go, C#
- Graph Algorithms - PageRank, BFS, shortest path, community detection
- Configuration - Server settings and tuning
GraphRAG and AI Integration
For developers building AI applications with graph knowledge:
- GraphRAG SDK - Core SDK for graph-based RAG
- LangChain Integration - LangChain graph tools
- LlamaIndex Integration - LlamaIndex graph index
- AG2 Integration - AutoGen agent framework support
Ecosystem Integration
Client Libraries
- All official clients (Python, JavaScript, Java, Rust, Go, C#)
GenAI & Agent Frameworks
Data Integration
APIs & Protocols
Deployment Platforms
Operations & Observability
Community and Support
- Discord: Join community for technical discussions
- GitHub Discussions: Ask questions
- Issue Tracker: Report bugs
- Documentation: Browse docs
License
FalkorDB is licensed under the Server Side Public License v1 (SSPLv1).