Back to all tools
Qdrant
High-performance Rust-based vector database with advanced filtering and sparse vectors.
0Open Source
Vector Database
Overview
Qdrant is a high-performance open-source vector database and similarity search engine written in Rust, with advanced filtering, sparse vectors, and a managed cloud offering.
Key Features
- Written in Rust — extremely high throughput and low latency
- Sparse + dense vector hybrid search
- Advanced payload filtering with complex conditions
- On-disk indexing for large datasets beyond RAM
- Multitenancy with collections and named vectors
- gRPC and REST API with Python, TypeScript, Go clients
Real-World Workflows
High-throughput recommendation system
- 1Embed user and item data into vectors
- 2Upload to Qdrant with rich metadata payloads
- 3Use payload filters (category, price, availability) with vector search
- 4Serve recommendations via gRPC for lowest latency
Getting Started
docker run -p 6333:6333 qdrant/qdrant pip install qdrant-client from qdrant_client import QdrantClient client = QdrantClient(host='localhost', port=6333) # Create collection and upsert vectors client.upsert(collection_name='docs', points=[...])
Compare Alternatives
See how Qdrant stacks up against similar tools.