Qdrant

Qdrant is a vector database engine built in Rust and optimised for production vector search. The platform powers retrieval-augmented generation (RAG) systems, semantic search, and embedding-based similarity applications. Unlike traditional databases, Qdrant indexes vectors for fast approximate nearest neighbor search at scale. Key features include hybrid search combining dense vectors with sparse vectors for improved retrieval quality, multivector support for multimodal AI applications (text + images + video), and advanced metadata filtering using JSON predicates with support for nested structures, text search, and geo-spatial queries. Real-time indexing avoids the costly full rebuilds required by some vector databases. Quantization techniques reduce memory footprint by up to 64x, making large-scale deployments more economical. Qdrant operates across multiple deployment scenarios: Qdrant Cloud for managed hosting on AWS, GCP, or Azure; Hybrid Cloud for Kubernetes deployments with decoupled control and data planes; Private Cloud for air-gapped or compliance-sensitive environments; and Edge (beta) for lightweight local search. The core engine is open-source (30,000+ GitHub stars) with commercial support available. Qdrant differentiates through sophisticated search capabilities, performance at scale, and deployment flexibility versus narrower competitors.