OSS Alternative - Discover Top Open Source Alternatives to Popular Software

ruvnet/RuVector

RuVector is a high-performance, self-learning vector database with integrated graph intelligence and local AI capabilities, designed for real-time, adaptive search and agentic systems.

Core Features

Self-learning and self-optimizing search quality via GNNs and SONA engine.

Hybrid search (sparse + dense vectors) and Graph RAG for superior retrieval.

Local LLM execution on hardware (CPU/GPU) without cloud API costs.

Full Cypher graph query engine for complex relationship analysis.

High-performance features like DiskANN, TurboQuant, and Matryoshka embeddings.

Quick Start

npx ruvector

Detailed Introduction

RuVector redefines the vector database paradigm by integrating self-learning, graph intelligence, and local AI directly into its core. Unlike traditional vector databases, it continuously optimizes search results and system performance based on usage patterns, leveraging Graph Neural Networks. It supports advanced retrieval techniques like hybrid search and Graph RAG, and enables cost-effective local execution of large language models. With a full Cypher query engine and high-performance optimizations, RuVector is ideal for building intelligent, adaptive, and agentic AI systems that learn and evolve.