What is pgvector 0.5.0 (HNSW Indexing)?

AI applications often turn data such as text into long lists of numbers called vectors, then need to find the stored items most similar to a query. Comparing against every item can be slow for large collections, so vector databases typically offer approximate nearest neighbor search, which checks only a subset of the data and can trade some result quality for speed. pgvector is an open-source extension that adds vector similarity search to the Postgres database; version 0.5.0 (2023-08-28) added HNSW, an index that links similar vectors into increasingly dense layers, and gives a better trade-off between search speed and result quality than IVFFlat, another index type in pgvector, but builds more slowly and uses more memory.

Written from primary sources and reference pages, and checked against them by an independent AI fact-checker. Drafted with AI assistance.