HelixDB

Active
GitHub Rust Apache-2.0

Description

An open-source graph-vector database built from scratch in Rust, combining graph database and vector retrieval capabilities to provide AI agents with unified storage for both knowledge graphs and semantic search.

Key Features

  • Graph-vector database combining graph traversal and vector similarity search in one system
  • Rust-based engine for high performance with low resource consumption
  • TypeScript and Rust SDKs with declarative query DSL for building queries
  • helix chef one-shot bootstrapper that scaffolds projects and hands off to coding agents
  • HelixDB Cloud with ACID transactions, auto-scaling readers, and high availability
  • Supports KV, document, and relational data models alongside graph and vector

Use Cases

💡 Building AI agent memory systems with both graph relationships and semantic vector search
💡 Creating knowledge graphs with federated data access for enterprise AI applications
💡 Running RAG pipelines that combine structured graph queries with unstructured vector retrieval
💡 Prototyping AI applications with a single database instead of managing multiple storage systems

Strengths & Limitations

Strengths

  • Actively maintained, recent updates
  • High community interest (5.7k stars)
  • Permissive open-source license (Apache-2.0)
  • Responsive to issues, low backlog

Quick Start

curl -sSL https://install.helix-db.com | bash && helix chef

Related Projects

CozoDB

4.1k · Rust
Stale C

CozoDB is a transactional, relational-graph-vector database that uses Datalog for queries. Designed as the hippocampus for AI, it unifies graph traversal, vector search, and relational queries.

vector-databasegraph-databasedatalog +2
  • · Datalog query language: powerful recursive queries with composable rules, superior to SQL for graph operations
  • · Embeddable database: runs in-process like SQLite or in client-server mode for higher concurrency
  • · Graph-native: handles graph traversal and algorithms (PageRank, community detection) natively within Datalog

HippoRAG

3.9k · Python
Active A+

NeurIPS 2024 RAG framework inspired by human long-term memory, combining knowledge graphs with personalized PageRank for continuous knowledge integration in LLMs.

ragknowledge-graphlong-term-memory +2
  • · Neurobiologically inspired RAG framework modeled after human hippocampal memory system
  • · Knowledge graph construction with personalized PageRank for associative multi-hop retrieval
  • · Continuous non-parametric learning that integrates new knowledge without retraining

FalkorDB

5.6k · Rust
Active B

A high-performance graph database built on GraphBLAS, optimized for LLM and GraphRAG scenarios with real-time knowledge graph construction and querying for graph-structured AI agent retrieval.

graph-databasegraphragknowledge-graph +1
  • · Graph database built on sparse matrices using linear algebra for efficient querying
  • · OpenCypher query language support with proprietary extensions
  • · Property Graph model with attributes on both nodes and relationships

Graphiti

30.0k · Python
Active A+

Graphiti is a temporal knowledge-graph engine for agent memory, helping systems continuously accumulate long-term context.

memoryknowledge-graphrag +1
  • · Temporal fact management — Each fact has a validity window; old facts are invalidated, not deleted. Query what's true now or at any point in time
  • · Provenance & lineage tracking — Every entity and relationship traces back to raw data episodes, full lineage from derived fact to source
  • · Hybrid retrieval — Combines semantic embeddings, keyword BM25, and graph traversal for low-latency high-precision queries without LLM summarization