Infinity
ActiveDescription
Infinity is an AI-native database providing incredibly fast hybrid search of dense vectors, sparse vectors, tensors, and full-text, designed for LLM applications and RAG systems.
Key Features
- Achieves 0.1ms query latency and 15K+ QPS on million-scale vector datasets
- Hybrid search across dense vectors, sparse vectors, tensors, and full-text
- Multiple built-in rerankers: RRF, weighted sum, and ColBERT
- Rich data type support: strings, numerics, vectors, and more
- Single-binary architecture with no dependencies, Docker one-click deployment
- Embeddable as a Python module, friendly to AI developers
Use Cases
💡 High-performance vector retrieval backend for RAG systems
💡 Core database for semantic search and recommendation systems
💡 Context retrieval and memory storage for conversational AI
💡 Hybrid retrieval and ranking for multimodal content
💡 Full-text and vector search fusion for large-scale document collections
Categories
Quick Start
Pull the Docker image and start the Infinity service: docker run -d --name infinity --network=host infiniflow/infinity:nightly, then install the Python client with pip install infinity-sdk to connect and use.