Pgvector 16

Pgvector is an extension for the Postgres database that allows it to store and search complex data, such as images or text, based on their meaning.

  • Databases
Pgvector is an extension for the Postgres database that allows it to store and search complex data, such as images or text, based on their meaning. It enables apps to offer “smart” features like personalised recommendations and more human-like search results by finding items that are similar.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

75/100 Good

Maintained
100last commit 3 days ago
Popular
8823,728 GitHub stars
Easy to install
150 install platforms, install notes
Light to run
No data
Security check
69no serious findings, CI, OpenSSF Scorecard 4.4/10

Based on 4 of 5 factors.

GitHub stars
23,728
Open issues
19
Last commit
2026-11-18
Activity
rising

Checked 3 days ago - source: GitHub (pgvector/pgvector)

Resources & Compatibility

Requirements not published yet.

First-install notes

  • Match your pgvector package to Postgres On Debian or Ubuntu, install the pgvector package for your Postgres version, for example sudo apt install postgresql-16-pgvector. The big number in the package name must match your Postgres server, or your database will not be able to load the extension. source
  • Add Postgres dev files before building If you build pgvector from source, install the Postgres development files first, for example sudo apt install postgresql-server-dev-16. These files include Postgres header files that the build needs; without them, make usually stops with a 'postgres.h not found' error. source
  • Use sudo only for make install When building from source, run make and then make install, using sudo for make install if your account is not allowed to write to the system Postgres folders. If the files are not placed where Postgres looks, the extension install will fail or the extension will still be missing. source

Fetched 53 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Why can’t PostgreSQL 16 load the pgvector extension after installing the package?

Use the pgvector package whose version number matches the Postgres server, for example postgresql-16-pgvector for Postgres 16. If the numbers do not match, the database cannot load the extension.

Why does building pgvector from source fail with 'postgres.h not found'?

Install the matching Postgres development files first, such as postgresql-server-dev-16 for Postgres 16. They provide the header files needed for the build.

How much RAM should I allow for pgvector’s HNSW index?

Plan for about 8 GB in practice, because the HNSW index wants to be resident in memory.

Can pgvector be used as the vector store in a self-hosted RAG app?

A self-hosted retrieval-augmented chatbot can use pgvector for the vector schema, embedding pipeline, and semantic search. It can also support a streaming answer endpoint.

Answers sourced from github.com, www.digitalapplied.com, www.macrostack.net

Community

Collected 48 days ago - public community threads