Zilliz는 위의 기능 외에도 모든 기능을 갖춘 클라우드 네이티브 경험을 제공합니다. Zilliz를 사용하면 Milvus 클러스터를 쉽게 관리하고 확장하며, 유사성 검색을 위해 벡터를 저장하고 검색하며, 컴퓨팅 리소스를 늘리지 않고도 모든 규모에서 비용 효율적으로 데이터를 유지할 수 있습니다.
한눈에 보기
Open Source
✅
❌
License
Apache 2.0
Proprietary
GitHub Stars
43,000+
N/A
Deployment
Docker, Helm, Kubernetes, SaaS
Cloud (fully managed)
Billion-Scale Vector Support
✅
✅
Search Engine
Open-sourced Knowhere
Cardinal (10x+ performance boost)
Open Source
✅
License
Apache 2.0
GitHub Stars
43,000+
Deployment
Docker, Helm, Kubernetes, SaaS
Billion-Scale Vector Support
✅
Search Engine
Open-sourced Knowhere
Open Source
❌
License
Proprietary
GitHub Stars
N/A
Deployment
Cloud (fully managed)
Billion-Scale Vector Support
✅
Search Engine
Cardinal (10x+ performance boost)
검색 기능
Vector Similarity Search
Efficiently store and search vectors for similarity. Fine-tuned parameters and software/hardware adaption for better performance.
✅
✅
Hybrid Search
Query multiple vector fields for multimodal, sparse-dense, and dense-text search for more accurate results.
✅ Multimodal, sparse-dense, dense-text
✅
Rich Schema Support
Easily work with a variety of data types, ie. float vector, integer, floating point, boolean, varchar, etc., for flexible schema modeling.
✅ Float vector, int, bool, varchar, etc.
✅
Auto ANNS Index
No more suffering on matching your workload to the right index type, and the endless optimizations. With AUTOINDEX, you can just focus on the business, and Zilliz Cloud will handle the rest.
❌ Manual index selection required
✅ AUTOINDEX — no manual tuning needed
Compatible SDKs
✅ Python, Java, JavaScript, Go, and Node.js
✅ Python, Java, JavaScript, Go, and Node.js
Built-in Embedding Pipelines
Convert unstructured data into searchable vector embeddings, seamlessly navigating from data preparation to chunking, model selection, and transformation.
Cluster types for different types of use cases.
✅ From raw data to searchable vectors — chunking, model selection, transformation
Vector Similarity Search
Efficiently store and search vectors for similarity. Fine-tuned parameters and software/hardware adaption for better performance.
✅
Hybrid Search
Query multiple vector fields for multimodal, sparse-dense, and dense-text search for more accurate results.
✅ Multimodal, sparse-dense, dense-text
Rich Schema Support
Easily work with a variety of data types, ie. float vector, integer, floating point, boolean, varchar, etc., for flexible schema modeling.
✅ Float vector, int, bool, varchar, etc.
Auto ANNS Index
No more suffering on matching your workload to the right index type, and the endless optimizations. With AUTOINDEX, you can just focus on the business, and Zilliz Cloud will handle the rest.
❌ Manual index selection required
Compatible SDKs
✅ Python, Java, JavaScript, Go, and Node.js
Built-in Embedding Pipelines
Convert unstructured data into searchable vector embeddings, seamlessly navigating from data preparation to chunking, model selection, and transformation.
Cluster types for different types of use cases.
Vector Similarity Search
Efficiently store and search vectors for similarity. Fine-tuned parameters and software/hardware adaption for better performance.
✅
Hybrid Search
Query multiple vector fields for multimodal, sparse-dense, and dense-text search for more accurate results.
✅
Rich Schema Support
Easily work with a variety of data types, ie. float vector, integer, floating point, boolean, varchar, etc., for flexible schema modeling.
✅
Auto ANNS Index
No more suffering on matching your workload to the right index type, and the endless optimizations. With AUTOINDEX, you can just focus on the business, and Zilliz Cloud will handle the rest.
✅ AUTOINDEX — no manual tuning needed
Compatible SDKs
✅ Python, Java, JavaScript, Go, and Node.js
Built-in Embedding Pipelines
Convert unstructured data into searchable vector embeddings, seamlessly navigating from data preparation to chunking, model selection, and transformation.
✅ From raw data to searchable vectors — chunking, model selection, transformation
운영 및 확장성
Fully Managed
Milvus clusters that are fully managed and automated with zero operations required.
❌ Self-managed
✅ Zero-ops, fully automated
High Availability
99.95% uptime SLA with built-in failover to ensure your Milvus clusters are always available.
❌ Manual setup
✅ 99.95% uptime SLA, built-in failover
Elastic Scaling
Easily scale up to a billion-scale or down without the need to over-provision infrastructure.
❌ Manual provisioning
✅ Scale to billions of vectors without over-provisioning
Machine Type Options
Provides two compute unit (CU) types: performance-optimized CU and capacity-optimized CU to fit different use cases.
❌
✅ Performance-optimized Capacity-optimized Tiered-storage For more details, check out this pricing page.
Infinite Storage
Cost-effectively store data at any scale without the need to increase compute resources.
User-managed storage scaling
✅ Fully managed storage-compute separation and scaling
Cloud UI
User-friendly GUI to easily manage and monitor your Milvus clusters at any scale in the cloud.
❌
✅ GUI to manage and monitor clusters at scale
Resource Monitors & Alerts
Get automatic notifications to avoid service overloading.
❌
✅ Automatic notifications to avoid service overloading
Fully Managed
Milvus clusters that are fully managed and automated with zero operations required.
❌ Self-managed
High Availability
99.95% uptime SLA with built-in failover to ensure your Milvus clusters are always available.
❌ Manual setup
Elastic Scaling
Easily scale up to a billion-scale or down without the need to over-provision infrastructure.
❌ Manual provisioning
Machine Type Options
Provides two compute unit (CU) types: performance-optimized CU and capacity-optimized CU to fit different use cases.
❌
Infinite Storage
Cost-effectively store data at any scale without the need to increase compute resources.
User-managed storage scaling
Cloud UI
User-friendly GUI to easily manage and monitor your Milvus clusters at any scale in the cloud.
❌
Resource Monitors & Alerts
Get automatic notifications to avoid service overloading.
❌
Fully Managed
Milvus clusters that are fully managed and automated with zero operations required.
✅ Zero-ops, fully automated
High Availability
99.95% uptime SLA with built-in failover to ensure your Milvus clusters are always available.
✅ 99.95% uptime SLA, built-in failover
Elastic Scaling
Easily scale up to a billion-scale or down without the need to over-provision infrastructure.
✅ Scale to billions of vectors without over-provisioning
Machine Type Options
Provides two compute unit (CU) types: performance-optimized CU and capacity-optimized CU to fit different use cases.
✅ Performance-optimized Capacity-optimized Tiered-storage For more details, check out this pricing page.
Infinite Storage
Cost-effectively store data at any scale without the need to increase compute resources.
✅ Fully managed storage-compute separation and scaling
Cloud UI
User-friendly GUI to easily manage and monitor your Milvus clusters at any scale in the cloud.
✅ GUI to manage and monitor clusters at scale
Resource Monitors & Alerts
Get automatic notifications to avoid service overloading.
✅ Automatic notifications to avoid service overloading
보안 및 규정 준수
24x7x365 Expert Support
Access to the world's foremost Milvus experts to offer you 24x7x365 support.
❌
✅
Compliance
SoC 2 Type II, ISO27001, GDPR-ready & HIPAA Compliant, providing robust data protection and privacy assurance.
❌
✅ SOC 2 Type II, ISO 27001, GDPR, HIPAA For more details, check out Zilliz trust center.
Data Recovery
Backup and restore, time travel and recycler bin to help restore data in the event of accidental loss.
✅ Backup & restore
✅ Backup, time travel & recycler bin
Data Encryption in Transit
Ensure the security of your data while it's being transferred.
✅
✅
Role-Based Access Control (RBAC)
Permission management to protect private and sensitive information and control user access.
✅ Basic RBACBasic RBAC
✅ Granular RBAC for fine-grained data protection
Bring Your Own Cloud (BYOC)
Opt-in BYOC for higher data security and compliance.
❌
✅ Deploy Zilliz Cloud inside your own cloud account (AWS, GCP, Azure available)
Cross-Region Disaster Recovery
❌
✅ For more details, check out Zilliz global cluster
24x7x365 Expert Support
Access to the world's foremost Milvus experts to offer you 24x7x365 support.
❌
Compliance
SoC 2 Type II, ISO27001, GDPR-ready & HIPAA Compliant, providing robust data protection and privacy assurance.
❌
Data Recovery
Backup and restore, time travel and recycler bin to help restore data in the event of accidental loss.
✅ Backup & restore
Data Encryption in Transit
Ensure the security of your data while it's being transferred.
✅
Role-Based Access Control (RBAC)
Permission management to protect private and sensitive information and control user access.
✅ Basic RBACBasic RBAC
Bring Your Own Cloud (BYOC)
Opt-in BYOC for higher data security and compliance.
❌
Cross-Region Disaster Recovery
❌
24x7x365 Expert Support
Access to the world's foremost Milvus experts to offer you 24x7x365 support.
✅
Compliance
SoC 2 Type II, ISO27001, GDPR-ready & HIPAA Compliant, providing robust data protection and privacy assurance.
✅ SOC 2 Type II, ISO 27001, GDPR, HIPAA For more details, check out Zilliz trust center.
Data Recovery
Backup and restore, time travel and recycler bin to help restore data in the event of accidental loss.
✅ Backup, time travel & recycler bin
Data Encryption in Transit
Ensure the security of your data while it's being transferred.
✅
Role-Based Access Control (RBAC)
Permission management to protect private and sensitive information and control user access.
✅ Granular RBAC for fine-grained data protection
Bring Your Own Cloud (BYOC)
Opt-in BYOC for higher data security and compliance.
✅ Deploy Zilliz Cloud inside your own cloud account (AWS, GCP, Azure available)
Cross-Region Disaster Recovery
✅ For more details, check out Zilliz global cluster