Documentation

Apache HugeGraph Documentation

Apache HugeGraph includes graph database, graph computing, and graph AI components. The HugeGraph core engine manages property graphs, transactions, and real-time queries; Computer and Vermeer run graph algorithms; and HugeGraph-AI provides GraphRAG, graph machine learning, and a Python client.

I want to…	Start here
Run graph queries (OLTP)	HugeGraph Server Quickstart
Large-scale graph computing (OLAP)	Graph Computing Engine
Build Graph + AI applications	HugeGraph-AI
Batch import data	HugeGraph Loader
Visualize and manage graphs	Hubble Web UI

Ecosystem Overview

┌─────────────────────────────────────────────────────────────────┐
│                  Apache HugeGraph Ecosystem                      │
├─────────────────────────────────────────────────────────────────┤
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────────────────┐  │
│  │ HugeGraph   │  │ HugeGraph   │  │ HugeGraph-AI            │  │
│  │ Core Engine │  │ Computer    │  │ (GraphRAG/ML/Python)    │  │
│  │ (OLTP)      │  │ (OLAP)      │  │                         │  │
│  └─────────────┘  └─────────────┘  └─────────────────────────┘  │
│         │               │                    │                   │
│  ┌──────┴───────────────┴────────────────────┴──────────────┐   │
│  │              HugeGraph Toolchain                          │   │
│  │  Hubble (UI) | Loader | Client (Java/Go/Py) | Tools      │   │
│  └───────────────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────────────┘

Core Components

HugeGraph Core Engine (OLTP): Exposes REST APIs through HugeGraph Server and supports Gremlin and Cypher queries
HugeGraph Toolchain: Includes Java/Go clients, Loader, Hubble, Spark Connector, and Tools; the Python client is maintained in HugeGraph-AI, and a Rust client is under development
HugeGraph Computer: Contains the distributed Computer engine and the in-memory Vermeer engine
HugeGraph-AI: Includes GraphRAG, graph machine learning, the Python client, and the Vermeer Python client

Deployment Modes

Mode	Core Components	Suitable Scenarios	Data Scale
Standalone	Server + RocksDB	Development, testing, and small to medium-scale data	≤ 2 TB
Distributed	Server + PD + Store (HStore)	Production, horizontal scaling, and multi-replica deployment	≤ 1 PB

See the system introduction and the corresponding quick-start guides for each component’s scope and startup instructions.

Get Started

Install HugeGraph, run the server, and make your first graph query.

Components

Start with HugeGraph, then explore toolchain, computing, and AI components.

Develop

Start with client APIs, then browse query languages and development guides.

Operate

Start with configuration, then browse security, backup, and performance guidance.

Reference

Start with release notes, then browse downloads, contribution, and legal references.