Weaviate

Weaviate is an open-source vector database from Weaviate B.V. in Amsterdam, led by co-founder and chief executive Bob van Luijt. It stores objects together with their vectors, so applications can combine semantic search with structured filters. The server is written in Go and published under the BSD 3-Clause licence, while enterprise features are covered by a commercial licence.

Features

• Hybrid search that combines vector search with BM25 keyword search, plus image search and advanced filtering
• Automatic vectorisation at import through integrated vectorizers, or import of embeddings that already exist
• Built-in generative search for retrieval-augmented generation and reranking
• Multi-tenancy, replication, role-based access control and vector compression, operated self-hosted or as a managed cloud service

Good to know

Weaviate sees itself as infrastructure for AI applications. It is heavier to run than a plain search library and contains more built-in AI functions than a pure vector index such as Qdrant. The company also has a US presence, and the Dutch B.V. is the contracting entity named in its privacy policy.