Embedded Graph Database
v0.1.0

GraphScope NeuG

High-performance embedded graph database for analytics and real-time transactions

Get StartedView on GitHub
terminal
$pip install neug
$python -c "import neug; print('Ready!')"
Ready!

Why NeuG?

A graph database that gives you the best of both worlds

Lightweight & Embeddable

Single binary, minimal external dependencies. Embed directly into your Python app for offline analytics, or run as a service for online transactions — no DevOps overhead.

Cypher-Native, GQL-Ready

Write queries in industry-standard Cypher. Powered by GOpt's unified IR design — ready for ISO/GQL with minimal migration cost.

ISO/GQL Standard →Learn about GOpt →

Extensible by Design

Postgres/DuckDB-inspired extension system. Keep the core lean. Add graph algorithms, vector search, or custom procedures through an extensible framework.

Proven Performance

80,000+ QPS

Built on GraphScope Flex, which set the record on LDBC SNB Interactive benchmark using Cypher queries.

View Official Audit Results
#1
LDBC SNB SF300

One Database, Two Modes

Switch between embedded analytics and network service based on your needs

Embedded Mode

For Analytics

Import as a library. Perfect for data science workflows, ML/AI pipelines, and research prototyping.

  • Batch data loading & ETL
  • Complex pattern matching
  • Full-graph algorithms

Use cases: Jupyter notebooks, offline analytics, graph ML feature extraction

Service Mode

For Transactions

Run as a network service. Built for production apps with concurrent users and real-time requirements.

  • Concurrent read-write
  • Low-latency point queries
  • Multi-session ACID transactions

Use cases: Web/mobile backends, real-time recommendations, anti-fraud systems

What's Next

v0.2 Roadmap

NeuG is actively evolving. Here's what we're working on:

Node.js Binding
AI Agent integration ready
Leiden Algorithm
Community detection for AI
Data Lake Support
S3/OSS + Parquet
Vector DB Extension
RAG & GraphRAG support

⭐ Star us on GitHub to stay updated on new releases

Get Started in Seconds

Install NeuG and create your first graph database

python
# Install
$pip install neug
# Create a graph database
import neug
db = neug.Database("") # in-memory
conn = db.connect()
# Run a Cypher query
result = conn.execute("MATCH (n) RETURN count(n)")
Python 3.8+
Linux & macOS
x86 & ARM
Read the Documentation

Ready to build with NeuG?

Open source, Apache 2.0 licensed. Start building your graph applications today.

Get StartedGitHub