The Faiss Library
Vector databases typically manage large collections of embedding vectors. As AI applications are growing rapidly, the number of embeddings that need to be stored and indexed is increasing. The Faiss library is dedicated to vector similarity search, a core functionality of vector databases. Faiss is a toolkit of indexing methods and related primitives used to search, cluster, compress and transform vectors. This paper describes the trade-offs in vector search and the design principles of Faiss in terms of structure, approach to optimization and interfacing. We benchmark key features of the library and discuss a few selected use cases to highlight its broad applicability.
Authors
- María Lomelí (ORCID: https://orcid.org/0000-0002-6524-3615)
- Pierre-Emmanuel Mazaré
- Matthijs Douze
- Chengqi Deng
- Gergely Szilvasy
- Alexandr Guzhva
- Jeff Johnson (ORCID: https://orcid.org/0000-0003-2743-0521)
- Lucas Hosseini (ORCID: https://orcid.org/0000-0003-0922-0109)
- Hervé Jeǵou
Institutions
- Institute on Taxation and Economic Policy (US)
- Milieux environnementaux, transferts et interactions dans les hydrosystèmes et les sols (FR)
- Electronic Arts (United States) (US)
- Skipton General Hospital (GB)
- Skin Research Center (FR)
- Zillow Group (United States) (US)
- Fairchild Semiconductor (United States) (US)
Publication Details
- Journal
- IEEE Transactions on Big Data
- Published
- 2025-10-13
- DOI
- https://doi.org/10.1109/tbdata.2025.3618474
- Citations
- 245
- Primary Topic
- Renaissance Literature and Culture
- Type
- article
- Field-Weighted Citation Impact
- 9094.30