Technology & AI
What is Vector Database?
A vector DB is the database optimized to store and query embeddings fast — the technical heart of any RAG or semantic search application.
Options: pgvector (Postgres + extension, simple if you already use Postgres), Pinecone (managed, scales easily), Qdrant/Weaviate (open-source), Supabase Vector (Postgres with pgvector ready).
For starters, pgvector handles up to millions of vectors. For hundreds of millions with complex filters, dedicated vector DBs win on latency.
Frequently asked questions
pgvector or Pinecone?
pgvector if you already use Postgres and have medium volume. Pinecone if you want managed scale without infra ops.
Does vector DB replace Postgres?
No. Coexists — Postgres for structured data, vector DB for semantic search. pgvector unifies both in one.