Qdrant

What it does

Qdrant is a vector search engine built in Rust. It supports metadata filters, hybrid dense and sparse search, multivector retrieval, and reranking. You can run it as open source or use Qdrant Cloud, a managed service on AWS, GCP, or Azure.

Qdrant is a vector search engine for teams building AI retrieval. The homepage says it is built entirely in Rust with SIMD and a custom storage engine called Gridstore. It supports metadata filters, hybrid dense and sparse search, multivector retrieval, and reranking. You can run it as open source or use Qdrant Cloud, a managed service on AWS, GCP, or Azure. The pricing page lists a free tier that is free forever for testing and prototypes, with a single node cluster of 0.5 vCPU, 1GB RAM, and 4GB disk. Paid tiers are usage-based for production and a premium tier with a minimum spend for enterprises. The homepage lists Qdrant Edge as beta and Qdrant Serverless as coming soon, so those options may not be ready for production. The page also mentions SOC2 and HIPAA compliance and enterprise features like SSO and private networking. Before you commit, check how filtering and scaling behave with your own data and review the tier limits on the pricing page.