Milvus
Open-source vector database built for billion-scale search
Milvus is profiled here as a RAG Framework tool for engineering teams. Read about features, pricing, and how it compares to related options in the tools directory.
Description
Milvus is an open-source vector database created by Zilliz in 2019 and now a graduated project of the LF AI and Data Foundation. Its architecture separates compute from storage so indexing and querying scale independently to billions of vectors across distributed nodes. Multiple index types let teams trade recall against speed and memory, and Zilliz Cloud offers the same engine as a managed service. Backed by Zilliz and a large contributor base, it serves vector workloads at scale across recommendation, search, and retrieval systems. The lightweight Milvus Lite build lets developers prototype locally before moving to a distributed cluster.
Key Capabilities:
Vector similarity search scaling to billions of embeddings
Multiple index types including HNSW, IVF, and DiskANN
Hybrid search over dense and sparse vectors with scalar filtering
Distributed architecture with separated compute and storage
Multi-tenancy through collections, partitions, and databases
Apache 2.0 license with the lightweight Milvus Lite for local use
Alternative tools
- LanceDB
Embedded multimodal vector database on the Lance format
- Sentence Transformers
Python framework for dense text and image embeddings
- R2R
Production retrieval system with ingestion and an API
- Mem0
Long-term memory layer for AI agents and assistants
- Docling
Open-source document conversion built for RAG pipelines
- Unstructured
Turn raw documents into LLM-ready structured data