This topic describes the billable items and billing methods of Elastic Algorithm Service (EAS) of Platform for AI (PAI).
Billing overview
When you use EAS to deploy services, you may be charged for computing resources, system disks, and dedicated gateways.
Computing resources: includes public resources, dedicated resources, and Lingjun resources.
(Optional) System disks: provides free quotas, including 30 GB for each instance created by using public resources and 200 GB for each instance created by using dedicated resources. You are charged for additional system disks.
(Optional) Dedicated gateways: By default, a free shared gateway is used. If you require security isolation, access control, or custom domain names, you can purchase a dedicated gateway. To use a dedicated gateway, you must manually configure it.
EAS provides the following billing methods:
Pay-as-you-go: You are charged based on your actual usage. This billing method is suitable for uncertain and fluctuating demand scenarios.
Subscription: This billing method is more cost-effective. It is suitable for long-term stable business scenarios.
EAS provides Stable Diffusion web UI Serverless Edition and ComfyUI Serverless Edition that you can use to deploy services free of charge. You are charged only based on the inference duration during service calling.
If you use other Alibaba Cloud services, such as Elastic IP Address (EIP), Object Storage Service (OSS), and File Storage NAS (NAS), fees are generated.
Billing period
Pay-as-you-go
For pay-as-you-go resources, you must pay special attention to the start and end time of billing to prevent unexpected fee deduction. To prevent unnecessary costs, we recommend that you stop model services that you no longer use.
Resource | Start time of billing | End time of billing |
Public resources | The time when the model service enters the Running state. You are charged for the model service even if you do not call the model service. | The time when a model service stops running and releases resources. Important You must stop or delete a service instance. |
Dedicated resources | The time when the model service is created and enters the Running state. You are charged for the model service even if the model service is not deployed. |
Important An instance can enter the stopped state only due to overdue payments. You must delete the pay-as-you-go instance to stop billing. |
System disk | The time when you purchase the system disk. |
|
Dedicated gateway | The dedicated gateway is created. The dedicated gateway is billed even if it is not used. | The time when the dedicated gateway is deleted. |
Service scaling
If you scale out a model service, you are charged for new resources after the scale-out operation is complete.
If you scale in a model service, the released resources are no longer billed. You are charged only for the remaining resources.
Subscription
Resource | Start time of billing | End time of billing |
Dedicated resources | The next day after the purchase, at 00:00:00. | The expiration time. |
AI computing resources (Lingjun resources) | The next day after the purchase, at 00:00:00. | The expiration time. |
System disk | The time when you purchase the system disk. | The expiration time. |
Dedicated gateway | The time when you purchase the dedicated gateway. | The expiration time. |
Service inference (Serverless Edition)
The deployment of Serverless Edition services is free of charge. You are charged only if you call the services based on the inference duration.
For example, if you click Generate on the web UI page to generate an image and the process requires 10 seconds, the billing duration is 10 seconds.
Only Stable Diffusion web UI and ComfyUI support the deployment of Serverless Edition services. You can view the pricing details on the deployment page.
Usage notes
The pricing information in this topic is provided only for reference. The actual prices in the billing statement shall prevail.
Billable resources
The following table describes the billing methods of different resources.
Billable item | Billing method | Fee calculation | Unit price | Usage notes |
Public resources | Pay-as-you-go (specific resources required by the instance) |
| Fees are calculated per minute on a pay-as-you-go basis and rounded down to the minute. For more information about the hourly prices of vCPU and memory resources, see Resource-based configuration. To calculate the unit prices per minute, divide the prices listed in Table 1 by 60. |
|
Pay-as-you-go (specific instance type) |
| Unit prices may vary based on the instance type and the region. See the prices in the console. For information about the instance types supported by the public resource group, see Appendix: Public resource group instance types. |
| |
Dedicated resources | Pay-as-you-go |
| For information about the unit prices of pay-as-you-go instances, go to the EAS post-payment for dedicated machine buy page. |
|
Subscription |
| For information about the unit prices of subscription instances, go to the EAS pre-payment for dedicated machine buy page. | Specific resources may be unavailable in specific regions for a short period of time. | |
AI computing resources (Lingjun resources) | Subscription | Lingjun resources support the subscription billing method. You can deploy model services in EAS by using Lingjun resource quotas. For more information, see Billing of AI computing resources. | ||
System disk | Pay-as-you-go |
| To view the unit prices of system disks, go to the product overview page of Elastic Compute Service (ECS), click the Pricing tab, and then click Storage. The pricing information of Enterprise SSD (ESSD) PL1 is displayed in the System Cloud Disk Fee section. | N/A. |
Subscription |
| To view the unit prices of system disks, go to the product overview page of ECS, click the Pricing tab, and then click Storage. The pricing information of ESSD PL1 is displayed in the System Cloud Disk Fee section. | N/A. | |
Dedicated gateway | Pay-as-you-go |
| To view the unit prices of dedicated gateways, go to the EAS Dedicated Gateway Postpay buy page. | N/A. |
Subscription |
| To view the unit prices of subscription dedicated gateways, go to the EAS dedicated gateway prepay buy page. | N/A. | |
Service inference (Serverless Edition) | Pay-by-inference-duration |
| Only Stable Diffusion web UI and ComfyUI support the deployment of Serverless Edition services. You can view the pricing details on the deployment page. | N/A. |
Table 1 Resource-based configuration
The unit prices listed in the following table are provided only for reference. You can view the actual prices in the console or on the buy page of the service resource that you want to purchase.
Resource | Pricing |
CPU | USD 0.03 per vCPU-hour |
Memory | USD 0.004 per GB-hour |
Billing examples
The following examples are provided only for reference. The actual prices listed on the EAS buy page of the PAI console shall prevail.
Public resource group
Pay-as-you-go
Scenario
You used the public resource group to deploy a model service in the China (Hangzhou) region. You specified the public resources used for model deployment based on your business requirements.
The model service occupied 2 vCPUs and 8 GB of memory and started to run at 09:00:00 (UTC+8) on June 3, 2019.
You scaled in the model service and decreased the occupied resources to 1 vCPU and 4 GB of memory at 10:00:00 (UTC+8) on June 3, 2019.
You scaled out the model service and increased the occupied resources to 4 vCPUs and 16 GB of memory at 11:00:00 (UTC+8) on June 3, 2019.
The model service stopped running at 12:00:00 (UTC+8) on June 3, 2019.
Fee calculation
Bill amount = 2 × 0.03 + 8 × 0.004 + 1 × 0.03 + 4 × 0.004 + 4 × 0.03 + 16 × 0.004 = USD 0.322
Dedicated resource group
Subscription
Scenario
You subscribed to two NVIDIA T4 GPUs in the China (Hangzhou) region for three months. Each GPU provides 4 vCPUs and 15 GB of memory. The unit price of the GPU is USD 570 per month.
Fee calculation
Bill amount = 2 × 570 × 3 = USD 3,420
Pay-as-you-go
Scenario
You purchased two pay-as-you-go ecs.g6.6xlarge instances in the China (Hangzhou) region and used the instances for 45 minutes. Each instance provides 24 vCPUs and 96 GB of memory. The unit price of the instance is USD 1.02 per hour.
Fee calculation
Bill amount = 2 × (1.02/60) × 45 = USD 1.53
System disk
Subscription
Scenario
You purchased two subscription instances in the China (Hangzhou) region for a dedicated resource group. Each instance provides a system disk capacity of 300 GiB (free storage: 200 GiB). The subscription duration is three months.
Fee calculation
Bill amount = 2 × (300 - 200) × 0.153 × 3 = USD 91.8
Pay-as-you-go
Dedicated resource group
Scenario
You purchased two pay-as-you-go instances in the China (Hangzhou) region for a dedicated resource group and used them for 5 hours. Each instance provides a system disk capacity of 300 GiB (free storage: 200 GiB).
Fee calculation
Bill amount = 2 × (300 - 200) × 0.000319 × 5 = USD 0.319
Public resource group
Scenario
You purchased two pay-as-you-go instances in the public resource group and deployed them in the China (Hangzhou) region. You used the instances for 5 hours. Each instance provides a system disk capacity of 300 GiB (free storage: 30 GiB).
Fee calculation
Bill amount = 2 × (300 - 30) × 0.000319 × 5 = USD 0.8613