pgvector project

license

PostgreSQL(OSI)

source

public

core features withheld

no

pgvector is released under the PostgreSQL License, the Postgres project's own permissive terms, which is OSI-approved. The repository is public, active and builds the whole extension, and the README describes no paid tier, enterprise edition or separately licensed build beside it.

• https://raw.githubusercontent.com/pgvector/pgvector/master/LICENSE recorded 2026-09-30

  LICENSE body carries the PostgreSQL License notices: "Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group" and "Portions Copyright (c) 1994, The Regents of the University of California", followed by a permission grant to use, copy, modify and distribute without fee.

• https://raw.githubusercontent.com/pgvector/pgvector/master/README.md recorded 2026-09-30

  README opens as "Open-source vector similarity search for Postgres" and gives build steps with "git clone --branch v0.8.6 https://github.com/pgvector/pgvector.git" then make and make install; searched the body for enterprise, paid, license key and commercial: none.

The pgvector PyPI package is pgvector-python, a client helper for talking to the extension from Python, so its download count belongs to that library rather than to the extension itself. The extension ships through operating-system distribution packages, PGXN and managed-Postgres vendors, none of which publish a comparable count, leaving GitHub stars as the only available adoption signal.

• https://api.github.com/repos/pgvector/pgvector recorded 2026-08-18

  Repo metadata - stargazers_count = 22,664 - for pgvector/pgvector.

pgvector adds exact and approximate nearest-neighbor search to Postgres, over four vector types and six distance metrics, using Postgres's own storage, transactions and durability rather than building them itself. It relies on Postgres for the rest of the retrieval system, which keeps it two steps below Vespa, a complete distributed retrieval system.

• https://raw.githubusercontent.com/pgvector/pgvector/master/README.md recorded 2026-09-30

  README lists "exact and approximate nearest neighbor search" over "single-precision, half-precision, binary, and sparse vectors" with "L2 distance, inner product, cosine distance, L1 distance, Hamming distance, and Jaccard distance", plus "compliance, point-in-time recovery, JOINs, and all of the other" Postgres features.

• https://raw.githubusercontent.com/pgvector/pgvector/master/README.md recorded 2026-09-30

  README describes indexing: "An HNSW index creates a multilayer graph." and "An IVFFlat index divides vectors into lists, and then searches a subset of those lists that are closest to the query vector."