This topic describes the billing rules of the computing power provided by Container Compute Service (ACS) and the billing rules of cloud services required by ACS.
ACS fees
ACS container compute power fee
Billing description
You are charged for pods based on the total resource limits of all containers within the pod (the cumulative sum of the CPU and memory limits for all containers in the pod). If the resulting vCPU and memory specifications are not supported by ACS, ACS automatically rounds up the specifications to the nearest predefined supported specification. For example, a specification of 2 vCPUs and 3.5 GiB of memory is rounded up to 2 vCPUs and 4 GiB of memory. You are charged based on the rounded-up specifications. For information about supported specifications, see Resource specifications.
Billing formula
Fee per pod = (Number of vCPUs × vCPU unit price + Memory size × Memory unit price) × Billing duration.
Fee per GPU-accelerated pod = (Number of vCPUs × vCPU unit price + Memory size × Memory unit price + Number of GPUs × GPU unit price) × Billing duration.
For detailed pricing, see Billing method.
Metering method
Pay-as-you-go fees are charged on a per-second basis by pod. Bills are generated on an hourly basis.
General-purpose pod billing
Region | Pod type | QoS class | Architecture | Billable item (billed on a per-second basis) | |
vCPU | Memory (GiB) | ||||
Chinese mainland | General-purpose | default | Intel | 0.0000026 USD/second (USD 0.00936/hour) | 0.00000128 USD/second (USD 0.0046/hour) |
best-effort | 0.00000052 USD/second (USD 0.001872/hour) | 0.00000025 USD/second (USD 0.0009/hour) | |||
Performance-enhanced | default | 0.0000034 USD/second (USD 0.01224/hour) | 0.00000167 USD/second (USD 0.006/hour) | ||
best-effort | USD 0.00000068/second (USD 0.002448/hour) | USD 0.000000334/second (USD 0.0012024/hour) | |||
China (Hong Kong) and regions outside China | General-purpose | default | 0.00000381 USD/second (USD 0.013716/hour) | 0.00000192 USD/second (USD 0.0069/hour) | |
best-effort | 0.00000076 USD/second (USD 0.002736/hour) | 0.00000038 USD/second (USD 0.001368/hour) | |||
Performance-enhanced | default | 0.00000496 USD/second (USD 0.017856/hour) | 0.00000248 USD/second (USD 0.008928/hour) | ||
best-effort | USD 0.000000992/second (USD 0.0035712/hour) | USD 0.000000496/second (USD 0.0017856/hour) | |||
For example, consider a general-purpose pod with the default QoS class in a Chinese mainland region. The pod has a configuration of 2 vCPUs and 8 GiB of memory. If billing starts at 10:05 and the bill is generated at 11:00, the pod runs for 3,300 seconds. The fee is calculated as follows: (2 vCPUs × USD 0.0000026/second + 8 GiB × USD 0.00000128/second) ×3,300 seconds ≈ USD 0.05.
Computing-accelerated pod billing
Region | Compute class | QoS class | Billable items (billed on a per-second basis) | |||
vCPU | Memory (GiB) | GPU | ||||
Chinese mainland | gpu | default | USD 0.0000026/second (USD 0.00936/hour) | USD 0.00000128/second (USD 0.004608/hour) | T4 (16 GB of GPU memory) | USD 0.00025889/second (USD 0.932/hour) |
A10 (24 GB of GPU memory) | USD 0.00028833/second (USD 1.038/hour) | |||||
best-effort | USD 0.00000104/second (USD 0.003744/hour) | USD 0.00000051/second (USD 0.0018432/hour) | T4 (16 GB of GPU memory) | USD 0.00010356/second (USD 0.3728/hour) | ||
A10 (24 GB of GPU memory) | USD 0.00011533/second (USD 0.4152/hour) | |||||
China (Hong Kong) and regions outside China | gpu | default | USD 0.00000381/second (USD 0.0137/hour) | USD 0.00000192/second (USD 0.0069/hour) | T4 (16 GB of GPU memory) | USD 0.00012222/second (USD 0.44/hour) |
A10 (24 GB of GPU memory) | USD 0.00039722/second (USD 1.43/hour) | |||||
best-effort | USD 0.00000152/second (USD 0.0055/hour) | USD 0.00000077/second (USD 0.0028/hour) | T4 (16 GB of GPU memory) | USD 0.00004889/second (USD 0.176/hour) | ||
A10 (24 GB of GPU memory) | USD 0.00015889/second (USD 0.572/hour) | |||||
For example, consider a GPU-accelerated pod with the default QoS class in a Chinese mainland region. The pod has a configuration of 8 vCPUs, 32 GiB of memory, and one T4 GPU. If billing starts at 10:05 and the bill is generated at 11:00, the pod runs for 3,300 seconds. The fee is calculated as follows: (8 vCPUs × USD 0.0000026/second + 32 GiB × USD 0.00000128/second + 1 GPU × USD 0.00025889/second) ×3,300 seconds ≈ USD 1.06.
Ephemeral storage billing
By default, a running ACS pod includes 30 GiB of free ephemeral storage (EphemeralStorage). If you need more storage, you can increase its size. Additional capacity is billed at the rates shown below.
Region | Billable items |
Chinese mainland | USD 0.0000000886/GiB/second (USD 0.00031896/GiB/hour) |
China (Hong Kong) and regions outside China | USD 0.0000000703/GiB/second (USD 0.00025308/GiB/hour) |
When a pod is hibernated, the free ephemeral storage tier does not apply, and you are billed for all of its storage capacity, including the initial 30 GiB. For more information, see Agent Sandbox billing.
Billing duration
The billing duration of an ACS pod begins when the image used by the pod starts being pulled and ends when the pod stops running. When the image is being pulled, the pod remains in the Pending state. When the pod stops running, the pod enters the Succeeded or Failed state. You do not need to pay for ACS pods that fail to be created or launched.
Billing of cloud services
If you use other Alibaba Cloud service resources when using ACS clusters and compute power, you are charged for those resources based on the billing rules of each cloud service, and the fees are collected by the respective cloud services.
Alibaba Cloud service | Type | Description | Subscription | Resource plan | Remarks |
VPC | Required | This service can be used to build networks and create routing rules for ACS clusters. For more information, see What is VPC | Not supported | Not supported | |
SLB | Recommended | This service allows you to enable load balancing for ACS clusters. For more information, see What is Classic Load Balancer (CLB)? and Application Load Balancer (ALB) |
|
|
|
Container Registry | Recommended | This service ensures the security of cloud-native applications that are fully managed on the cloud and allows you to manage the lifecycle of these applications. For more information, see What is Container Registry? | Supported | Not supported | |
Simple Log Service | Recommended | This service allows you to collect and query the log data of ACS cluster components and applications. For more information, see What is Simple Log Service? | Not supported | Supported | |
EIP | Optional | This service enables Internet access for cloud resources. For more information, see What is an EIP? | Not supported | Not supported | |
Managed Service for Prometheus | Required | Used for metrics collection, storage, and alerting for clusters and containers. | Not supported | Not supported | |
NAT Gateway | Optional | This service enables Internet access for ACS clusters and allows ACS clusters to pull images over the Internet. For more information, see What is NAT Gateway? | Not supported | Supported | |
NAS | Optional | This service allows you to store application data in NAS file systems. For more information, see What is Apsara File Storage NAS? | Not supported | Supported | |
Object Storage Service (OSS) | Optional | This service allows you to store application data in OSS buckets. For more information, see What is OSS? | Not supported | Supported | |
Key Management Service (KMS) | Optional | This service allows you to manage and encrypt Secrets in ACS clusters. For more information, see What is Key Management Service? | Not supported | Not supported | |
Alibaba Cloud DNS PrivateZone | Optional | Alibaba Cloud DNS PrivateZone is a DNS resolution service for private domain names within VPCs. For more information, see What is PrivateZone | Not supported | Not supported |
Container Registry Enterprise Edition supports only the subscription billing method. For more information, see Billing of Enterprise Edition instances.