The word "Voyager" in white on a black background.

Voyager is a library for performing fast approximate nearest-neighbor searches on an in-memory collection of vectors.

Voyager features bindings to both Python and Java, with feature parity and index compatibility between both languages. It uses the HNSW algorithm, based on the open-source hnswlib package, with numerous features added for convenience and speed. Voyager is used extensively in production at Spotify, and is queried hundreds of millions of times per day to power numerous user-facing features.

Think of Voyager like Sparkey, but for vector/embedding data; or like Annoy, but with much higher recall. It got its name because it searches through (embedding) space(s), much like the Voyager interstellar probes launched by NASA in 1977.

Python compatibility#

Voyager 2.2.0 requires Python 3.9 or newer. Python 3.8 is no longer supported.

Official support covers the Python versions and platforms in the compatibility table. Release wheels are tested on each listed combination before publication.

Standard CPython 3.12 and later share cp312-abi3 wheels. These wheels may also install on newer, untested Python versions; installation alone does not establish official support. New versions become officially supported when their wheel tests are added to CI and pass. Free-threaded Python builds are not currently supported.