Weaviate
By Weaviate
Weaviate is an open-source vector database that stores data objects together with their vector embeddings, enabling fast semantic and hybrid search over unstructured data such as text, images, and other media.
Definition
Weaviate is an open-source vector database that stores data objects together with their vector embeddings, enabling fast semantic and hybrid search over unstructured data such as text, images, and other media.
Overview
Vector databases emerged to support similarity search for AI applications, and Weaviate is one of the most widely adopted options in that category. It stores each data object alongside a vector embedding and supports built-in vectorization modules that connect directly to embedding model providers, so teams don't need to manage a separate embedding pipeline. Under the hood, Weaviate uses HNSW-based approximate nearest neighbor indexing for low-latency similarity search, and it supports hybrid search that blends dense vector similarity with traditional keyword (BM25) scoring for more reliable results. It exposes both GraphQL and REST APIs, and its modules can auto-generate embeddings using providers such as OpenAI or models hosted on Hugging Face. Weaviate sits at the center of most retrieval-augmented generation (RAG) pipelines, where it's frequently paired with orchestration frameworks like LangChain to fetch relevant context before it's passed to a language model. Learners building this kind of pipeline in the Retrieval-Augmented Generation course, or reading the Vector Databases Explained post, will encounter Weaviate as one of the standard building blocks of modern AI applications.
Key Features
- Native vector storage paired with structured object properties
- HNSW-based approximate nearest neighbor indexing for low-latency similarity search
- Hybrid search combining dense vector similarity with keyword (BM25) scoring
- Built-in and pluggable vectorizer modules for popular embedding providers
- GraphQL and REST APIs alongside official client libraries
- Multi-tenancy and horizontal scaling for production workloads
- Support for filtering, aggregations, and generative search (RAG) queries
Use Cases
History
Weaviate is an open-source, AI-native vector database for storing objects together with their vector embeddings, enabling semantic (similarity) search combined with structured filtering at scale. It was founded in 2019 in Amsterdam by Bob van Luijt (CEO) and Etienne Dilocker (CTO); the company was originally known as SeMI Technologies. Weaviate stood out early by letting developers query with GraphQL rather than a bespoke query language and by supporting rich filters and aggregations alongside vector search. Offered both self-hosted and as a managed cloud service, it became a widely used backbone for retrieval-augmented generation (RAG) and other large-language-model applications.
Sources
- Weaviate — "The History of Weaviate" · as of 2026-07-17
- Weaviate on GitHub — weaviate/weaviate · as of 2026-07-17