Qdrant is an open-source (Apache 2.0) vector database written in Rust providing high-performance similarity search with rich filtering capabilities. Open-source is free to self-host. Qdrant Cloud offers free tier (1GB), Starter at $25/month, and scaling tiers. The Rust implementation provides memory efficiency and performance advantages—important for cost-sensitive deployments. Advanced filtering enables combining vector search with arbitrary payload filters (scalar, keyword, geo, nested) with optimization for filtered queries. Quantization options (scalar, product, binary) reduce memory requirements up to 32x with minimal accuracy loss. Sparse vectors support enables hybrid search combining dense and sparse representations. Features include sharding, replication, and snapshot-based backups. For text-to-SQL: efficient storage of schema and query embeddings with complex filtering (by database, table, query type). Page should cover: filtering capabilities deep-dive, quantization configuration, sparse vectors setup, self-hosted deployment guide, comparison with Pinecone and Weaviate, pricing analysis, and optimization for database context retrieval.