PostgreSQL with pgvector

PostgreSQL with the pgvector extension is your unified vector database solution: store, index, and query high-dimensional embeddings directly within your existing Postgres infrastructure.

This is the smart way to handle vector data: use the PostgreSQL extension, pgvector. It adds a dedicated `vector` data type, supporting up to 2,000 dimensions, and integrates vector similarity search right into your relational database. You gain efficiency by eliminating a separate vector database, simplifying your stack. For performance, it offers both exact and approximate nearest neighbor search, leveraging specialized indexes like HNSW and IVFFlat. Need to find the closest matches for an OpenAI embedding or run a semantic search? Use standard SQL with operators for L2 distance, inner product, or cosine similarity. It's a powerful, consolidated platform for your modern AI applications.

https://github.com/pgvector/pgvector