Qdrant is a high-performance vector database written in Rust, known for its strong metadata filtering capabilities and excellent performance-per-dollar. Available as open source for self-hosting or as managed cloud.
Qdrant stores vectors in HNSW (Hierarchical Navigable Small World) graphs for fast approximate nearest neighbor search. Metadata is indexed separately and can be used for pre-filtering queries, which Qdrant handles particularly well — many vector DBs struggle when you combine filtering with high-dimensional search. Multi-tenancy is first-class via collections and payload-based isolation.
Open source: free. Qdrant Cloud: Free tier (1GB), Paid tiers from $0.05/hour. Enterprise on-prem available.
Deloitte, Flipkart, HubSpot, Bosch, Bayer, and many others. Popular with teams that need strong filtering combined with vector search.