Q.What is a Compute Unit (CU)?
A.A Compute Unit (CU) is a group of hardware resources used to serve indexes and search requests, acting as a fully managed physical node for deploying search services.
Z
Zilliz
★4.3
💬16652
💲Freemium
Zilliz Cloud offers a fully managed vector database for enterprise AI applications. It enables fast, scalable, and secure similarity searches across images, text, audio, and other data types. Built on Milvus, it supports Retrieval Augmented Generation (RAG), large language models, and AI-driven search use cases with high performance and availability.
💻
Platform
web
AICloud ComputingCloud DatabaseData ManagementEnterprise AIHigh PerformanceLarge Language Models
What is Zilliz?
Zilliz is a fully managed, scalable vector database designed for enterprise AI applications. Built on the open-source Milvus platform, it supports large-scale vector similarity searches, Retrieval Augmented Generation (RAG), and integration with large language models. It simplifies the deployment and management of AI-driven search applications by removing infrastructure complexity, making it ideal for developers and data scientists working with AI and machine learning.
Core Technologies
- Vector Database
- Artificial Intelligence
- Machine Learning
- Retrieval Augmented Generation (RAG)
- Large Language Models
- Cloud Database
- Open Source
- Data Management
- High Performance
- Enterprise AI
Key Capabilities
- Billion-scale vector search
- Fully managed Milvus service
- Multi-cloud availability
- Built-in embedding pipelines
- High availability and performance
- Security and compliance standards
- Comprehensive data management
- Observability and role-based access control
Use Cases
- Retrieval Augmented Generation (RAG)
- Recommender systems
- Text and semantic search
- Image similarity search
- Audio similarity search
- Video similarity search
- AI agents
- Molecular similarity search
- Multimodal similarity search
Core Benefits
- Reduces operational overhead with fully managed service
- Supports billion-scale vector search for AI applications
- Offers flexible pricing with a free tier and pay-as-you-go
- Provides robust security and compliance standards
- Enables seamless integration with major cloud providers
- Includes built-in tools for data preparation and observability
- Scales easily for high QPS and large datasets
Key Features
- Fully managed Milvus service
- Billion-scale vector search
- High performance with Cardinal search engine
- Highly scalable up to 500 CUs
- High availability with 99.95% uptime
- Security and governance with SOC2 and ISO27001
- Built-in embedding pipelines
- Multi-cloud availability
- AI integrations
- Comprehensive data management
- Observability with metrics and alerts
- Role-based access control
How to Use
- 1Sign up for a free account
- 2Download an official SDK (Python, Java, Go, Node.js)
- 3Create your first collection
- 4Perform vector similarity searches
- 5Upgrade to a paid plan for production use
Pricing Plans
Free
$0/mo.
A starting point for learning, experimenting, and prototyping, with easy migration to paid plans. Includes 5 GB storage (enough for 1M 768 dim vectors), 2.5M vCUs per month, and up to 5 collections. Serverless.
Serverless
From $0.3 /GB per month
Pay only for what you use. Auto-scaling. Up to 100 collections. For applications with variable or infrequent traffic. Minimal configuration required.
Dedicated
From $99 /mo.
Dedicated clusters offer use case optimized CUs to achieve high control, consistent performance, and cost-effectiveness. Suitable for development and testing. Includes multiple cloud providers and regions, use case optimized CU types, and basic metrics and monitors. Up to 30-day free trial available.
BYOC (Bring Your Own Cloud)
Contact Us
Designed for organizations prioritizing custom infrastructure, enhanced data protection, and compliance. Deploy on your infra of choice with enhanced data control and security, and flexibility and scalability on demand.
Frequently Asked Questions
Q.What is a vCU?
A.A virtual Compute Unit (vCU) measures resource consumption for read and write operations like search, query, insert, and delete.
Q.Which type of CU should I pick?
A.Choose Performance-optimized CU for real-time applications, Capacity-optimized for large datasets, or Extended-capacity CU for massive-scale data with cost optimization.
Q.How many CUs do I need for a given collection?
A.Performance-optimized CU supports up to 1.5M vectors, Capacity-optimized up to 5M, and Extended-capacity up to 20M 768-dimensional vectors.
Q.How can I get Zilliz Cloud discounts?
A.Commit to an annual plan to receive additional usage-based credits.
Q.How can I request a new cloud region?
A.Fill out the form on the Zilliz website to request a new cloud service provider region.
Pros & Cons (Reserved)
✓ Pros
- Fully managed service reduces operational overhead
- Exceptional performance and scalability for billion-scale vector search
- Multi-cloud support across AWS, Azure, and GCP
- Robust security and compliance standards
- Flexible pricing options including a free tier
- Backed by the popular open-source Milvus project
- Comprehensive features for data management and observability
- Built-in embedding pipelines simplify data preparation
✗ Cons
- Optimizing for specific use cases can be complex due to multiple CU types and pricing models
- Requires a foundational understanding of vector databases and AI concepts for advanced configuration
- Cost can scale significantly for very large datasets and high QPS requirements
Alternatives
No alternatives found.