pgvector

Overview

pgvector adds vector similarity search to PostgreSQL. Store embeddings alongside your regular data — no separate vector database, no data sync, no new infrastructure. Use your existing Postgres for semantic search, RAG, recommendations, and deduplication. Supports exact and approximate nearest neighbor search with IVFFlat and HNSW indexes.

When to Use

• Adding semantic/vector search to an existing Postgres-backed app
• RAG (Retrieval-Augmented Generation) without running Pinecone/Qdrant/Weaviate
• Storing embeddings alongside relational data (users, products, documents)
• Recommendation systems based on content similarity
• Don't want to manage a separate vector database

Instructions

Setup