SoftwareFramework

Faiss

Project record

Maintained by Meta (Fundamental AI Research)2 5 3

Faiss is a library for similarity search and clustering of dense vectors, written in C++ with Python wrappers. It includes exact and approximate nearest-neighbor indexes, among them compressed (quantization-based) and graph-based (HNSW, NSG) indexes, and GPU implementations of several algorithms. It is developed mainly at Meta's Fundamental AI Research (FAIR) group, which released it in March 2017.2 5 8Fact reviewed Oct 1, 2026

Repository: Repository (external site: github.com)
Documentation: Documentation (external site: faiss.ai)
Documentation: Wiki (tutorial and FAQ) (external site: github.com)
License: LICENSE (MIT) (external site: github.com)
Release notes: Releases (external site: github.com)

Last reviewedOct 1, 2026Entry updated Oct 1, 2026Documented release Mar 2017

Availability and license

Overall availability

Public

Source code on GitHub under the MIT License. Prebuilt packages are distributed through the pytorch conda channel.1 3 4

Availability fact review: Oct 1, 2026

Availability is separate from permission: read the license before using or redistributing.

MIT License (external site: github.com)3
Applies to: Code · SPDX MIT · on the rubric’s OSI-approved software-license list · fact reviewed Oct 1, 2026

Component reuse rights

A readable or downloadable component is not automatically reusable. These indicators concern recorded license evidence, not system certification.

weights: Unknown — no complete fact-level rights review
code: Reviewed qualifying license recorded — check scope and conditions
data: Unknown — no complete fact-level rights review
documentation: Unknown — no complete fact-level rights review

No complete system-rights review is recorded for this release.

Public materials checklist

Items for a framework under USASI rubric v0.2. Unknown means unassessed or insufficient evidence.

Public materials checklist for Faiss
Item	Status	Notes and evidence
Source codeIs the source code publicly readable?	Public	Public GitHub repository.1
DocumentationIs user documentation published?	Public	A GitHub wiki with a tutorial, FAQ, and troubleshooting section, plus per-class C++ API documentation at faiss.ai.2 5 6
InstallationAre installation instructions or packages publicly available?	Public	INSTALL.md names conda as the supported install path (faiss-cpu, faiss-gpu, and faiss-gpu-cuvs packages, plus nightly builds) and also documents Pixi and building from source with CMake.4
Supported platformsAre supported operating systems or hardware documented?	Public	faiss-cpu conda packages cover Linux (x86-64 and aarch64), macOS (arm64), and Windows (x86-64); faiss-gpu covers Linux x86-64 with CUDA 11.4 and 12.1, and faiss-gpu-cuvs covers Linux x86-64 with CUDA 13.2. The README describes optional GPU support through CUDA or AMD ROCm; INSTALL.md says ROCm conda packages are not yet available.4 2
Release statusAre versioned releases published?	Public	Versioned releases are published on GitHub; the latest at review was v1.15.1, published September 16, 2026.7

What it is useful for

Searching large collections of vectors by L2 distance, dot product, or cosine similarity, for example the embeddings used in retrieval systems, including collections that do not fit in RAM; also k-means clustering and tools for tuning index parameters.2 5Fact reviewed Oct 1, 2026

Run and use notes

Documented facts only. No hardware or performance claims are made without a cited source and stated assumptions.

The README says the only required dependency is a BLAS implementation; the GPU code and the Python interface are optional, and NVIDIA cuVS GPU backends can be enabled.2

Organization context

Meta

U.S. eligibility

Project eligibility rests on documented governing or maintaining entities, not on contributors.

Eligible · basis: U.S.-governed project

The README says Faiss is developed primarily at Meta's Fundamental AI Research group and carries a Meta Platforms, Inc. copyright notice, and the MIT license names Facebook, Inc. and its affiliates as copyright holder. Meta Platforms, Inc. is a Delaware corporation with principal executive offices in Menlo Park, California, per the cover page of its fiscal 2025 Form 10-K.2 3 9

Assessed Oct 1, 2026

Sources

1.
facebookresearch/faiss (GitHub repository) (external site: github.com)
Meta · Repository · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
2.
facebookresearch/faiss README.md (external site: raw.githubusercontent.com)
Meta · Repository · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
3.
facebookresearch/faiss LICENSE (MIT) (external site: raw.githubusercontent.com)
Meta · License · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
4.
Installing Faiss (INSTALL.md) (external site: raw.githubusercontent.com)
Meta · Documentation · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
5.
Welcome to Faiss Documentation (external site: faiss.ai)
Meta · Documentation · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
6.
Getting started · facebookresearch/faiss Wiki (external site: github.com)
Meta · Documentation · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
7.
Faiss release v1.15.1 (external site: github.com)
Meta · Release notes · published Sep 16, 2026 · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
8.
Faiss: A library for efficient similarity search (external site: engineering.fb.com)
Engineering at Meta · Announcement · published Mar 29, 2017 · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026
9.
Meta Platforms, Inc. Form 10-K for fiscal 2025, cover page (XBRL viewer) (external site: sec.gov)
U.S. Securities and Exchange Commission (filed by Meta Platforms, Inc.) · Filing · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

Email a correction Report a correction Edit this entry

This listing is not an endorsement, a safety assessment, or a federal approval.