Extension that adds vector search to PostgreSQL.
Category: Memory & Data · Type: Postgres vector extension · Website: https://github.com/pgvector/pgvector
Overview
pgvector adds vector data types and similarity search to PostgreSQL, letting teams add RAG to an existing Postgres database without adopting a separate vector store. It's the pragmatic choice for teams already on Postgres.
Key Features
- Vector similarity search inside Postgres
- Keeps relational data + embeddings together
- Supported by major Postgres clouds
- No new infrastructure to operate
Role in the Agentic AI Stack
The lowest-friction vector option for Postgres-based stacks; an alternative to standalone stores like Qdrant and Milvus.
Part of Agentic AI Landscape → Memory & Data