# Milvus

> An open-source vector database built for GenAI applications with high-speed searches and scalability to tens of billions of vectors.

Milvus is an open-source vector database designed specifically for GenAI applications, enabling developers to build powerful AI-driven solutions with high-performance vector similarity search capabilities. It supports scaling to tens of billions of vectors with minimal performance loss, making it ideal for machine learning, deep learning, similarity search tasks, and recommendation systems.

- **Elastic Scalability** allows the database to scale horizontally to handle billions of vectors with a fully distributed architecture, supporting datasets from prototyping to enterprise-grade production workloads.

- **Blazing Fast Retrieval** leverages Global Index technology to retrieve data quickly and accurately regardless of scale, ensuring optimal performance for real-time AI applications.

- **Multiple Deployment Options** include Milvus Lite for learning and prototyping via pip install, Milvus Standalone for single-machine production deployments with up to millions of vectors, Milvus Distributed for enterprise-grade scalability, and Zilliz Cloud for fully managed vector database service.

- **Zilliz Cloud Managed Service** provides a fully managed Milvus experience with 10x faster performance, offering Standard, Enterprise, and Business Critical tiers with features like serverless scaling, dedicated clusters, enterprise security controls, and HIPAA-eligible compliance options.

- **Feature-Rich Capabilities** encompass metadata filtering, hybrid search, multi-vector support, RAG (Retrieval-Augmented Generation), image search, multimodal search, and Graph RAG functionality.

- **Extensive Integration Support** works seamlessly with popular AI development tools including LangChain, LlamaIndex, Haystack, DSPy, FastGPT, Dify, and many other orchestration frameworks.

- **Developer-Friendly Tools** include Attu for GUI management, Milvus CLI for command-line operations, sizing tools for capacity planning, backup utilities, and Vector Transport Service (VTS) for data migration.

- **Reusable Code Architecture** enables developers to write once and deploy with minimal code changes into production environments, streamlining the development workflow.

To get started with Milvus, install it via pip using `pip install pymilvus` for Milvus Lite, deploy using Docker or Kubernetes for larger-scale implementations, or sign up for Zilliz Cloud for a fully managed experience. The quickstart guide provides code examples for creating collections and performing vector searches in just a few lines of Python code. The active community offers support through Discord, GitHub discussions, and regular office hours.

## Features
- Vector similarity search
- Elastic horizontal scaling
- Global Index for fast retrieval
- Metadata filtering
- Hybrid search
- Multi-vector support
- RAG (Retrieval-Augmented Generation)
- Image search
- Multimodal search
- Graph RAG
- Milvus Lite for prototyping
- Milvus Standalone deployment
- Milvus Distributed for enterprise
- Zilliz Cloud managed service
- Python SDK (PyMilvus)
- Attu GUI management
- Milvus CLI
- Backup and restore tools
- Vector Transport Service (VTS)

## Integrations
LangChain, LlamaIndex, Haystack, DSPy, FastGPT, Dify, Langflow, DocsGPT, PrivateGPT, Dynamiq, Llama Stack, n8n, AnythingLLM, Kotaemon, OpenAI, Zilliz Cloud

## Platforms
WINDOWS, MACOS, LINUX, WEB, API, DEVELOPER_SDK

## Pricing
Open Source, Free tier available

## Version
2.6.x

## Links
- Website: https://milvus.io
- Documentation: https://milvus.io/docs
- Repository: https://github.com/milvus-io/milvus
- EveryDev.ai: https://www.everydev.ai/tools/milvus