qdrant

Why we included this project

Qdrant is a vector database written in Rust that stores embeddings alongside JSON payload metadata, so each point can carry the business attributes you need to filter on. Its extended filtering support is the main reason teams pick it for production: you can combine approximate nearest-neighbor search with precise conditions on the payload in a single query, instead of retrieving candidates and filtering them in application code. That makes it a solid fit for semantic search and retrieval-augmented features where results have to respect real-world criteria. It runs as a self-hosted service with REST and gRPC APIs and official client libraries, and the same engine is available as a managed cloud offering, so moving from a local prototype to a deployed system is fairly direct. If you are choosing a vector store for a real product, Qdrant is worth benchmarking against your own data and query patterns.