About pgvector
pgvector adds vector similarity search to PostgreSQL. It lets you store embeddings next to your normal tables and search them with SQL.
What it does
pgvector is an extension you install into a Postgres database. It supports exact and approximate nearest neighbor search, and it works with the rest of Postgres features such as joins and transactions. Developers commonly use it for semantic search, recommendations and AI features that depend on embeddings.
Key features
- Exact and approximate nearest neighbor search
- Several vector types, including half precision and sparse
- Distance functions such as L2, cosine and inner product
- Works with joins and ACID transactions
- Quantization for large vector sets
Who it's for
Backend developers and small teams that already run Postgres and want vector search without adding a separate database. AI app builders often start here.
Pricing and good to know
pgvector is open source and free to use. It can be installed from source on Linux and Mac, or through Docker, Homebrew and several package managers. It needs PostgreSQL 13 or newer, and many hosted Postgres providers already include it.
Highlights
- Exact and approximate nearest neighbor search
- Several vector types, including half precision and sparse
- Distance functions such as L2, cosine and inner product
- Works with joins and ACID transactions
- Quantization for large vector sets
