Qdrant

Qdrant develops a high-performance vector search engine and database designed for AI retrieval, similarity matching, and semantic search at scale. Built entirely in Rust with SIMD optimization and a custom storage engine called Gridstore, Qdrant delivers fast, scalable vector search without wrappers or bolt-ons. The engine supports native hybrid search combining dense and sparse vectors with BM25, SPLADE++, and miniCOIL support, advanced metadata filtering with nested, text, geo, and has_vector filters applied during HNSW traversal for one-stage filtering with high recall and low latency. Qdrant also offers built-in multivector support for multimodal retrieval, full-spectrum reranking with score boosting, late interaction models like ColBERT, and Maximum Marginal Relevance for result diversification. The platform provides real-time indexing without rebuilding, memory-efficient storage for billions of vectors, and asymmetric, scalar, and binary quantization that reduces memory usage by up to 64x while maintaining search quality. Qdrant serves use cases including RAG and GenAI, AI agents with persistent memory, semantic search, recommendation systems, and data analysis and anomaly detection across industries like e-commerce, legal tech, hospitality, HR tech, and healthcare. The platform offers multiple deployment models: Qdrant Cloud on AWS, GCP, or Azure with auto-sharding; Qdrant Hybrid Cloud for bring-your-own Kubernetes; Qdrant Private Cloud for air-gapped deployments; and Qdrant Edge for low-latency search near data sources. With over 30,000 GitHub stars and 60,000 community members, Qdrant is SOC2 and HIPAA compliant. Founded in 2020 and headquartered in Berlin, Germany, Qdrant has raised approximately $87.8 million in Series B funding from investors including Spark Capital, Unusual Ventures, Bosch, and IBB Ventures.