SemaDB
A multi-index, multi-vector, document-based vector database and search engine with a JSON RESTful API, combining vector, keyword, geo and filter search in a single query.
SemaDB is a vector database and search engine built around a document-based, multi-index, multi-vector model with a JSON RESTful API. It originated as the search components of a knowledge-management project at Semafind before becoming a standalone open-source project, and the maintainers describe it as stable and running in production for several years.
It performs vector search using the graph-based Vamana algorithm for approximate nearest-neighbor search, alongside keyword and phrase search, geo indices based on latitude and longitude or geohashes, multi-vector search across several vectors per document, quantized vector representations to reduce memory, and hybrid search that combines vector and keyword results with adjustable weights. All of these search types, plus metadata filters, can be combined in a single query, and the whole database ships as one binary with no additional services to install.
It is licensed under Apache-2.0.