Qdrant
ActiveDescription
Qdrant is a high-performance vector database widely used as the retrieval layer for RAG and agent memory search scenarios.
Key Features
- Dense/Sparse/Multi-vector Search — Semantic similarity, full-text search, and multi-embedding retrieval like ColBERT
- Payload Filtering — Attach JSON payloads to vectors with keyword, full-text, numeric range, and geo-location filters
- Qdrant Edge — Lightweight edge deployment version for offline use with Qdrant server synchronization
- Multi-language Clients — Official clients for Go, Rust, JavaScript/TypeScript, Python, .NET, and Java
- REST & gRPC Interfaces — OpenAPI 3.0 REST API and high-performance gRPC interface
- Cloud & Self-hosted — Managed Qdrant Cloud with free tier, or deploy with Docker and bare metal
Use Cases
💡 Provide vector retrieval layer for RAG applications with semantic search and knowledge-base Q&A
💡 Build recommendation systems with similar item matching based on user behavior and content embeddings
💡 Implement multimodal search such as image similarity search and cross-modal retrieval
💡 Supply long-term memory storage for AI agents with context retrieval and experience replay
💡 Build product categorization and extreme classification systems in e-commerce scenarios
Tags
Categories
Quick Start
docker run -p 6333:6333 qdrant/qdrant
# Python client connection
pip install qdrant-client
from qdrant_client import QdrantClient
client = QdrantClient(url="http://localhost:6333")