NebulaGraph
NebulaGraph is a high-performance, distributed, native graph database designed for large-scale graphs with millisecond response times. Designed by the team at, it enables fast, efficient, multi-hop searches on graphs with hundreds of billions of nodes and trillions of edges.
NebulaGraph’s unique architecture separates the compute and storage engines allowing for unlimited scaling of both workloads across multiple machines. With it’s automated sharding feature and a “Shared-Nothing” storage approach, NebulaGraph is well suited for applications requiring high availability. Although the graph database product is built for maximizing transaction processing performance, NebulaGraph implements a Hybrid Transactional / Analytical Processing database model with its companion Analytics engine that enables rapid graph traversal and data analysis. In addition, NebulaGraph offers full native support for the Graph Query Language standard including GQL-specific architecture optimizations and specialized graph algorithms.
NebulaGraph is offered as a fully managed Cloud solution or as an on-premise solution. It adopts the Apache 2.0 license and comes with a wide range of data visualization tools.
History
NebulaGraph was developed in 2018 by Vesoft Inc. In May 2019, NebulaGraph made its free software available on GitHub, and its alpha version was released in the same year.In June 2020, NebulaGraph raised $8M in a series pre-A funding round led by Redpoint China Ventures and Matrix Partners China.
In June 2019, NebulaGraph 1.0 GA was released, while version 2.0 GA was released in March 2021. The latest open-source version 3.8.0 of NebulaGraph was released in May 2024.
In September 2023, NebulaGraph and LlamaIndex introduced Graph RAG for retrieval-augmented generation.
In July 2024, NebulaGraph Enterprise V5.0 was released introducing a redesigned architecture and the industry’s first Graph Query Language compatible graph database.
In November 2025, NebulaGraph Enterprise V5.2 was released. A major upgrade built for Graph Intelligence with a lightweight in-database compute engine, 100x faster path queries, and native graph-vector-text hybrid retrieval.