Compute & scaling
- Multiple instance families and sizes
- Usage-based or hourly billing
- API and console provisioning
- Scaling options for changing demand
Official-source provider dossier
Alibaba Cloud
Raw cloud infrastructure · CLI / unmanaged
Alibaba Cloud Elastic Compute Service is an IaaS platform with x86, Arm, GPU, bare-metal and high-performance instance families across global regions.
Raw cloud infrastructureHosting modelCLI / unmanagedManagementPlan-specificCheck the catalog belowNot benchmarkedReason documented below
Shared plan catalog
Inspect a specific plan.
These are the same records used by the finder and comparison. Research notes below describe the provider, not a guarantee for every plan.
Loading plans…
At a glance
What Alibaba Cloud offers
Best for APAC-focused and enterprise workloads needing Alibaba Cloud ecosystem integration and flexible instance billing.
Billing methodsSubscription / pay as you go / Spot
Pay-as-you-go meterPer secondMinimum duration by vCPU count
Spot discountUp to 90%Provider-published
Architecturesx86 + Arm
Independent benchmarkNot yet publishedNo controlled HostingHunt test
Provider feature research
Features arranged by category
Security
- Private networking and identity controls
- Disk/image security options
- Customer configures workload-level security
- Compliance and availability vary by region/service
Cloud operations
- Images, disks and network resources are selected separately
- Monitoring and automation integrations
- Snapshots and backups are separate resources
- Stopping compute may not stop every attached-resource charge
Pricing model
- Region and instance family affect price
- Operating system licenses may add cost
- Outbound transfer and public IP charges may apply
- Calculator or live configuration is required for comparison
Platform coverage
Infrastructure & supported workloads
Infrastructure options
Applications and use cases
Decision guide
Strengths and considerations
Why it may fit
- Broad instance catalog
- Strong APAC footprint
- Multiple commitment models
- Terraform and native orchestration support
Check before buying
- Every region has different unit prices
- Stopped instances can retain billable disks and IP resources
- Spot instances can be reclaimed
- Public bandwidth and snapshots can add cost