LanceDB

Multimodal lakehouse for AI.

LanceDB is an open-source multimodal lakehouse for AI designed to manage petabyte-scale data. It caters to a range of users:

1. Software developers and AI engineers

   As a vector retrieval and memory layer for large-scale Generative AI and search applications leveraging LLMs and embeddings — such as RAG, chatbots, agents, e-commerce search & recommendation systems, and more.

2. AI researchers and data scientists

   As a data management platform for large-scale AI workflows: model fine-tuning and training, feature engineering and explorative data analytics over petabyte-size datasets.

Products

LanceDB offers a range of products to suit different needs, listed below.

LanceDB Cloud

Sign up for LanceDB Cloud: Serverless vector search with enterprise-grade performance!

With LanceDB Cloud , you can build secure, scalable, and high-performance vector search solutions. Manage your Projects through the LanceDB Cloud UI , and connect via our simple REST API and SDKs in Python, Typescript and Rust .

LanceDB Enterprise

LanceDB Enterprise is both a private cloud or a BYOC solution that transforms your data lake into a high-performance vector database or lakehouse that can operate at extreme scale.

LanceDB Enterprise is designed for very large-scale workloads in private deployments, with high performance at extreme scale, or if you have strict security requirements, contact us to learn more.

LanceDB OSS

LanceDB is also available as an open source embedded database. This is the simplest way to get started with LanceDB, with client SDKs available in Python, TypeScript and Rust. You can use LanceDB OSS to build search applications, RAG apps and more by running it entirely locally on your filesystem.

Quickstart

Get started with LanceDB in minutes.

Build a RAG App

Learn how to build Retrieval-Augmented Generation (RAG) applications using LanceDB.