The release of Kùzu v0.13.6 marks another major milestone in the evolution of this project. This update introduces key enhancements, performance optimizations, and critical bug fixes that make building graph-powered applications, local retrieval-augmented generation (RAG) pipelines, and network analysis tools easier than ever.
Kùzu v0.1.3.6 refines this experience. By prioritizing core engine stability, optimizing memory bounds during graph traversal, and preserving its friction-free setup, it empowers software engineers and data scientists to build graph-powered applications with unparalleled agility. Whether you are constructing a complex GraphRAG agent, building a local recommendation engine, or looking to untangle messy relational data joins, Kùzu v0.1.3.6 provides a fast, light, and powerful framework to let your data connect. kuzu v0 136
The v0.13.6 release focuses on refining the database's storage engine, hardening memory limits during heavy analytical workloads, and expanding the language ecosystem integrations. 1. Enhanced Memory Hardening and Spill-to-Disk The release of Kùzu v0
: Now supports arbitrary Cypher queries for filtering vector search results, providing greater flexibility in data retrieval. Whether powering fraud detection networks
Graph databases have transitioned from niche tools for academic research into critical infrastructure for modern enterprises. Whether powering fraud detection networks, driving real-time recommendation engines, or serving as the knowledge retrieval backbone (GraphRAG) for Large Language Models (LLMs), the demand for fast, efficient graph data management has never been higher.
Leveraging the optimized execution engine of v0.13.6, you can cleanly query your graph:
The open-source community answered this dilemma with , an in-memory, embedded graph database management system (DBMS) designed for query speed and seamless integration. Built in C++, Kùzu brings the same philosophy to graph data that DuckDB brought to relational data: serverless simplicity, extreme efficiency, and native integration with modern analytical tools.
