Ask HN: Why would I use a vector db instead of the db I already use?
I use postgres and it has pgvector - why should I consider using Chroma, Pinecone, or any other dedicated vector db for my llm-based application that needs to store embeddings information? You can do everything with PostGres: Full-text search, but there are better engines for it: Elastic, Meilisearch, etc. right? You can also store JSON into Postgres, but you should better use MongoDB for NoSQL purposes, right? The reason for this is: dedicated tools are always better, faster, and more feature-rich. The ANN index IVF implemented in pgvector has very poor performance, with only around 50% recall. Is it something you are looking for? Disclaimer: I'm a co-founder at Qdrant, an open-source vector database written in Rust. https://github.com/qdrant/qdrant PS: Chroma isn't a database but а Python wrapper around ClickHouse DB.