tensorchord/pgvecto.rs
A Postgres extension written in Rust that provides vector similarity search for AI and LLM applications.

pgvecto.rs is a Postgres extension enabling vector similarity search within PostgreSQL. It is implemented in Rust using pgrx and supports high-dimensional vectors (up to 65535 dimensions) with dynamic SIMD dispatch for performance. The extension integrates with Postgres to provide hybrid-enabled vector search and filtering capabilities, making it suitable as a backend for RAG (Retrieval Augmented Generation) systems and LLM applications requiring semantic search.