⬢github C · 283 ★ +206 since we first saw it · pushed 1 d ago · PostgreSQL
timescale/pg_vectorsearch
High-performance vector search for PostgreSQL
pg_vectorsearch is a PostgreSQL extension (alpha) by Timescale for approximate nearest-neighbor search. Its PRISM index combines hierarchical clustering, RaBitQ quantization with error bounds, and SIMD-optimized distance computations to deliver faster queries and index builds than pgvector's IVFFlat and HNSW, at 100M+ vector scale, working with both native vector types and existing pgvector columns.
Why now: Recently announced with benchmarks claiming it outperforms pgvector's IVFFlat and HNSW at equivalent recall on 100M vectors with a quarter of the memory, which sparked Hacker News discussion.
Who it is for: Developers running vector similarity search in PostgreSQL, especially at large scale, who want faster ANN indexes without migrating off pgvector.
Stars over our 56 snapshots: 77 to 283, since 14 h ago.
Where people talked about it
API: https://socialmediatrends-api.osmike.com/v1/repos/timescale/pg_vectorsearch