faiss

Implement vector similarity search with FAISS and NumPy.

Updated May 9, 2026

Search Tags:#GPU acceleration #vector search #similarity search #FAISS #vector clustering

Ready to run

npx skills add https://github.com/robertbr123/Linket-Agent --skill faiss-robertbr123

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: faiss
Source: https://github.com/robertbr123/Linket-Agent/tree/main/optional-skills/mlops/faiss
Command: npx skills add https://github.com/robertbr123/Linket-Agent --skill faiss-robertbr123

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires faiss-cpu, faiss-gpu, numpy, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill solves the challenge of efficient similarity search and clustering of dense vectors, providing high-performance vector retrieval and pure similarity search without metadata.

Core Features & Use Cases

High-Performance Vector Search: Offers fast k-NN search, large-scale vector retrieval, and pure similarity search for high-performance applications.
GPU Acceleration: Supports GPU acceleration for processing billions of vectors, reducing search time significantly.
Multiple Index Types: Provides various index types including Flat, IVF, HNSW, and Product Quantization for different use cases.
Use Case: Ideal for scenarios requiring rapid vector search in applications like recommendation systems, content search, and real-time search.

Quick Start

Use the faiss skill to search for the 5 nearest neighbors of a given vector.

Frequently Asked Questions about faiss

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I perform fast vector similarity search on billions of dense vectors?▼

Vector similarity search on a massive scale requires optimized indexing and GPU acceleration. This Skill uses the FAISS library to provide high-performance k-NN retrieval, enabling rapid searching through billions of dense vectors using GPU processing.

How do I find the k-nearest neighbors for a given vector?▼

To find k-nearest neighbors, you query an indexed dataset of dense vectors. This Skill supports pure similarity search, allowing you to quickly retrieve the closest matching vectors for applications like content search and recommendation systems.

Can I use GPU acceleration for vector search and what are the prerequisites?▼

GPU acceleration for vector search is supported to significantly reduce search time. You need to have the FAISS GPU libraries installed alongside NumPy, or you can run the implementation on CPU using the FAISS CPU libraries.

What is the best way to handle large-scale vector clustering and retrieval?▼

Handling large-scale vector clustering and retrieval requires specialized index types. This Skill provides multiple FAISS index options including Flat, IVF, HNSW, and Product Quantization to optimize diverse high-performance search applications.

Does pure similarity search include filtering by metadata?▼

Pure similarity search focuses strictly on high-performance vector retrieval and does not include metadata filtering. It is designed for applications needing rapid dense vector matching without the overhead of metadata processing.

Why use FAISS for vector search over other similarity search libraries?▼

Using FAISS for vector search provides billion-scale processing capabilities through GPU acceleration. It supports multiple index types like HNSW and IVF, making it ideal for high-performance recommendation systems and real-time search applications.