kitio.aiAdded Oct 28, 2026
Faiss
FreeOpen-source library for similarity search over dense vectors
What is Faiss?
Faiss is a library for fast similarity search and clustering of dense vectors, written in C++ with Python wrappers. It handles collections of any size, including ones that do not fit in RAM, and offers GPU implementations of its most useful algorithms. It is developed mainly at Meta's FAIR research team.
You build an index from your vectors, then query it to get the nearest neighbor or the k nearest, and you can search several queries at a time. The library includes supporting code for evaluating results and tuning parameters.
It is a building block rather than a hosted product, commonly placed underneath retrieval and recommendation systems. Documentation includes a wiki and a full C++ API reference.
Features
- Nearest neighbor search
- Clustering
- GPU algorithms
- Python wrappers
- Indexes larger than RAM
- Evaluation tools
Tags
Releases
+ Submit new release- Listed
Faiss was added to the directory.
Organization
Meta FAIR
🇺🇸 United States
Listed by kitio.ai · unverified owner
Pricing
- Pricing model
- Free
- Paid options from
- 100% free
Sponsored
Your AI tool here
A Featured launch puts your tool in this spot and on the home page for a week.
Top Faiss alternatives
View allModels & APIs
Open-source engine for search, recommendation and RAG at scale
Models & APIs
The main hub for open models, datasets and AI demo apps
Models & APIs
Run open models locally with one command, plus optional cloud models
If you liked Faiss
High-throughput open-source engine for serving LLMs in production
FreeModels & APIsMade in United StatesRun and deploy open-source AI models through a simple API
PaidModels & APIsMade in United StatesFast inference and fine-tuning platform for open models
PaidModels & APIsMade in United StatesOpen-source deep learning library that runs on JAX, PyTorch and TF
FreeModels & APIsMade in United States
Pros and cons
Pros
- Free and open source
- Scales past available memory
- GPU support for key algorithms
Cons
- A library, not a managed service
- Needs engineering effort to deploy
- Not a database, so no built-in server or storage
Prompts & results
No prompts yet. Add your own prompt and output to help others use this tool.
Share a prompt
Reviews
Q&A
What is Faiss?
Faiss is a library for fast similarity search and clustering of dense vectors, written in C++ with Python wrappers. It handles collections of any size, including ones that do not fit in RAM, and offers GPU implementations of its most useful algorithms. It is developed mainly at Meta's FAIR research team.
Which languages can I use it from?
It is written in C++ and has complete wrappers for Python.
Is there a hosted version?
No. Faiss is a library you run yourself; hosted vector databases are separate products.
How much does Faiss cost?
Faiss is 100% free.
Is there a free version of Faiss?
Yes. Faiss is completely free to use.
What are the main features of Faiss?
Nearest neighbor search, Clustering, GPU algorithms, Python wrappers, Indexes larger than RAM, Evaluation tools.
Who makes Faiss?
Faiss is made by Meta FAIR.
What are the best alternatives to Faiss?
Popular alternatives include pgvector, Vespa, Hugging Face.