Qdrant

The product is a vector database for AI retrieval and similarity search on unstructured data, supporting real-time indexing, hybrid dense and sparse search, filtering, and reranking for use cases like RAG, recommendation, and AI agents. It is for developers, data teams, and IT teams in businesses building production AI applications. It is delivered as open-source software with managed cloud, hybrid cloud, private cloud, and edge deployment options.

Key features

• High-performance vector search at scale
• Real-time indexing without rebuild
• Memory-efficient storage for billions of vectors
• Asymmetric, scalar and binary quantization
• Expansive JSON metadata filtering
• Native hybrid dense and sparse search
• BM25, SPLADE++ and miniCOIL support
• Multivector per object support
• One-stage filtering during HNSW traversal
• Full-spectrum reranking with MMR
• Score boosting and ColBERT late interaction
• REST, gRPC and client APIs
• Built-in web UI and visualizations
• Native cloud inference for embeddings
• Rust with SIMD and Gridstore engine