OSS Alternative - Discover Top Open Source Alternatives to Popular Software

tensorchord/pgvecto.rs

A scalable, low-latency PostgreSQL extension written in Rust, enabling advanced vector similarity search directly within your relational database.

Core Features

Hybrid vector and relational queries using the VBASE method.

Supports high-dimensional vectors (up to 65535 dimensions).

Optimized performance with dynamic SIMD instruction dispatch.

Introduces new data types: binary vectors, FP16, and INT8.

Manages index storage and memory independently from PostgreSQL.

Quick Start

docker run --name pgvecto-rs-demo -e POSTGRES_PASSWORD=mysecretpassword -p 5432:5432 -d ghcr.io/tensorchord/pgvecto-rs:pg17-v0.4.0

Detailed Introduction

pgvecto.rs is a powerful PostgreSQL extension built with Rust, designed to integrate high-performance vector similarity search capabilities directly into your existing relational database. Unlike standalone vector databases, it allows users to perform complex hybrid queries combining vector search with traditional relational filtering and joins, leveraging the familiarity and robustness of PostgreSQL. It offers significant advantages over alternatives like pgvector, including support for higher dimensions, advanced data types, and optimized performance, aiming to revolutionize vector search without requiring a complete database migration.