Weaviate

Weaviate is an open-source vector database designed for building AI-powered applications that require semantic search, retrieval-augmented generation, and hybrid search capabilities. Founded in 2019 in Amsterdam, Weaviate stores data objects alongside their vector embeddings and supports fast similarity search at scale using approximate nearest neighbor (ANN) algorithms. Unlike pure vector stores, Weaviate is a full-featured database that supports structured filtering, CRUD operations, multi-tenancy, and ACID transactions alongside vector search, making it suitable for production applications that need both semantic understanding and traditional data management. A distinguishing feature of Weaviate is its built-in vectorization modules that can automatically generate embeddings using integrated models from OpenAI, Cohere, Hugging Face, Google, and others, eliminating the need for developers to manage a separate embedding pipeline. Weaviate also supports hybrid search that combines vector similarity with keyword-based BM25 scoring, improving retrieval accuracy for many use cases. The database offers a GraphQL API and REST API for querying and data management, along with client libraries for Python, JavaScript, Go, and Java. Weaviate supports generative search modules that pipe retrieval results directly into LLMs for augmented generation, making it a comprehensive solution for RAG applications. It can be deployed self-hosted using Docker or Kubernetes, or through Weaviate Cloud Services (WCS), the managed cloud offering. Weaviate Cloud offers a free sandbox tier for experimentation, a Serverless tier with pay-as-you-go pricing based on stored objects, and an Enterprise tier with dedicated resources and premium support. The open-source version under the BSD-3-Clause license can be self-hosted at no cost. Weaviate is used by companies and developers building search engines, recommendation systems, chatbots, and knowledge management applications.