Use pgvector for Vector Similarity Search
pgvector is an open-source plugin for vector similarity search. It supports both exact and approximate nearest neighbor searches, as well as L2 distance, inner product, and cosine distance. For more details, see pgvector/pgvector: Open-source vector similarity search for Postgres.
Apache Cloudberry allows you to use pgvector for data storage, querying, indexing, hybrid searches, and more through SQL statements. This document explains how to use pgvector in Apache Cloudberry.
While the pgvector extension is maintained by the dedicated pgvector community, Cloudberry is committed to ensuring its compatibility and ease of use within our platform. Here is one forked pgvector version, which is contributed by the community members and customized for Cloudberry, but please note that it is not maintained as one official Cloudberry project.
Quick start
Enable the extension (do this once in each database where you want to use it):
CREATE EXTENSION vector;