Qdrant

Qdrant is an open-source vector search engine and database written in Rust and licensed under Apache 2.0. It stores vectors alongside structured payload data and filters on that payload during search. The core engine ships as a single self-contained binary with REST and gRPC APIs, deployable with Docker, on Kubernetes through the official Helm chart and Operator, or on Qdrant's own managed cloud, so teams choose where the database runs.

Every part of retrieval, from indexing and scoring to filtering and ranking, is exposed as a composable primitive rather than a black box. A single query can mix dense vectors, sparse vectors, metadata filters, and custom scoring logic, which suits hybrid search and recommendation workloads that need fine-grained tuning. Companies such as Tripadvisor, HubSpot, and Canva run it in production.

Self-hosting is free with no licence fee. Qdrant Cloud adds a free-forever single-node cluster for testing, a usage-based Standard tier billed hourly on the resources a cluster uses, and a Premium tier with SSO, private VPC links, and a higher uptime SLA above a minimum spend. Hybrid Cloud lets Qdrant Cloud manage clusters running inside a customer's own Kubernetes, and Private Cloud covers fully isolated and air-gapped deployments for regulated workloads. Cloud Inference provides hosted embedding models on the managed service.