DocsDatabases

pgvector

A Postgres extension that adds a vector type, indexes and operators. Lets you reuse the database you already trust.

At a glance

License

PostgreSQL License

Hosting

Any Postgres ≥ 13

Indexes

IVFFlat, HNSW

Strength

Joins, transactions, RLS

Enable and create a table

sql

create extension if not exists vector;

create table docs (
  id      bigserial primary key,
  content text,
  embedding vector(1536)
);

create index on docs
  using hnsw (embedding vector_cosine_ops)
  with (m = 16, ef_construction = 64);

Query

sql

select id, content
from   docs
order  by embedding <=> $1   -- cosine distance
limit  5;

One database, fewer moving parts

For most teams under ~50M vectors, pgvector inside an existing Postgres beats running a separate vector service. You get joins, transactions, row-level security and your existing backups for free.