StreamingDiskANN index-supported queries

You need to use the cosine-distance embedding measure (<=>) in your ORDER BY clause. A canonical query would be:

Pgvector provides a graph-based indexing algorithm based on the popular HNSW algorithm.

To create an index named document_embedding_idx on table document_embedding having a vector column named embedding, run:

This command creates an index for cosine-distance queries because of vector_cosine_ops. There are also "ops" classes for Euclidean distance and negative inner product:

Distance type	Query operator	Index ops class
Cosine / Angular	<=>	vector_cosine_ops
Euclidean / L2	<->	vector_ip_ops
Negative inner product	<#>	vector_l2_ops

Pgvector HNSW also includes several index build-time and query-time parameters.