Zilliz

A fully managed, scalable vector database optimized for enterprise AI applications and large-scale machine learning projects.

Visit Site

AI Developer Tools AI Search Engine AI Knowledge Base Large Language Models (LLMs)AI For Data Analytics

About Zilliz

Zilliz Cloud offers a fully managed vector database powered by the open-source Milvus project. Designed for enterprise-grade AI applications, it supports billion-scale vector searches, Retrieval Augmented Generation (RAG), and large language models. The platform simplifies deploying and scaling vector search solutions by removing the complexity of infrastructure management, enabling faster innovation.

How to Use

Create a Zilliz Cloud account, access the platform via APIs or SDKs (Python, Java, Go, Node.js), and set up your first vector collection. Perform similarity searches and upgrade to paid plans as needed. The platform provides user-friendly REST APIs and SDKs for seamless control and data management, ideal for deploying AI-powered applications.

Features

Exceptional performance with 10x faster retrieval using Cardinal search technology

Fully managed Milvus vector database service

Role-based access control for secure data management

High availability with 99.95% uptime guarantees

Integrated AI features and embedding pipelines

Real-time observability with metrics, alerts, and monitoring tools

Supports billion-scale vector searches

Comprehensive data management including migration, import, and backup

Available across multiple cloud providers such as AWS, Azure, and GCP

Security standards including SOC2 Type II and ISO27001 compliance

Highly scalable architecture supporting up to 500 compute units and 100 billion items

Use Cases

Personalized Recommendation Systems

Multimodal Similarity Search

Image Similarity Analysis

AI Agent Deployment

Audio Content Similarity Search

Retrieval Augmented Generation (RAG) for AI

Video Content Similarity Search

Molecular Structure Similarity

Semantic and Text Search

Best For

AI and ML ResearchersAI Developers and Data ScientistsOrganizations Requiring Large-Scale Vector SearchEnterprise AI Application ArchitectsData EngineersTeams Working with Large Language Models

Pros

Supports multiple cloud providers including AWS, Azure, and GCP

Complies with industry security standards like SOC2, ISO27001, GDPR, and HIPAA

Delivers high performance and scalable vector search at billion-scale levels

Built-in embedding pipelines streamline data preparation

Comprehensive features for data management, monitoring, and access control

Fully managed service reduces operational complexity

Flexible pricing tiers, including free and pay-as-you-go options

Cons

Requires foundational knowledge of vector databases and AI concepts for advanced features

Complexity in optimizing configurations due to diverse compute unit types and pricing

Costs can escalate significantly with very large datasets and high query rates

Pricing Plans

Choose the perfect plan for your needs. All plans include 24/7 support and regular updates.

Free Tier

$0/month

Ideal for learning, testing, and prototyping, with easy upgrade options. Includes 5 GB storage (enough for 1 million 768-dimensional vectors), 2.5 million virtual compute units monthly, and up to 5 collections. Serverless setup for simplified use.

Get Started

Serverless Plan

From $0.3 per GB per month

Pay only for your actual usage. Features auto-scaling and support for up to 100 collections. Suitable for applications with fluctuating or low traffic, requiring minimal setup.

Get Started

Dedicated Clusters

Starting at $99 per month

Dedicated clusters with optimized compute units for high control and consistent performance. Ideal for development, testing, and production workloads. Supports multiple cloud providers and regions, with a 30-day free trial.

Get Started

Bring Your Own Cloud (BYOC)

Designed for organizations needing customized infrastructure with enhanced data security and compliance. Deploy on your preferred cloud environment with flexible scalability and control.

Get Started

Frequently Asked Questions

Find answers to common questions about Zilliz

What is a Compute Unit (CU)?

A compute unit (CU) is a set of hardware resources dedicated to hosting your indexes and managing search requests. It functions as a fully-managed physical node within the platform.

What does a vCU measure?

A virtual compute unit (vCU) measures resources used during read operations like search and query, as well as write operations such as insertions and deletions. Costs vary based on usage.

Which type of CU should I choose?

Choose performance-optimized CUs for real-time, high-concurrency search needs. Capacity-optimized CUs suit large datasets with reliable speed. Extended-capacity CUs are best for massive datasets prioritizing cost over latency.

How many CUs are needed for my collection?

Performance CUs support up to 1.5 million 768-dimensional vectors. Capacity CUs handle up to 5 million, and extended CUs support up to 20 million vectors. These figures are for vectors with primary keys only.

How can I get discounts on Zilliz Cloud?

Committing to annual plans can earn you additional credits based on your usage volume.

How do I request a new cloud region?

Fill out the regional expansion request form on our website to suggest new cloud provider locations.