All Products
Search
Document Center

Platform For AI:Billing of DLC

Last Updated:May 16, 2024

This topic describes the billing rules of Deep Learning Containers (DLC) of Platform for AI (PAI).

Billing

The following figure shows how DLC is billed.

fef886433e30d82cb1e606a927adb6ad

Billing methods

The following table describes the billing methods for public and dedicated resource groups.

Billable resource

Billable item

Billing rule

Billing method

How to stop billing

Public resource group

The running duration of a DLC job, which is the duration for which the public resource group is occupied.

The public resource group is billed based on the amount of time during which public resources are occupied by the DLC job.

Pay-as-you-go

Stop the DLC training job or wait until the job is completed.

Dedicated resource group

The running duration of computing resources in the dedicated resource group.

You are charged only for the computing resources in the dedicated resource group. You are not charged for DLC jobs that are deployed by using the dedicated resource group.

Subscription

N/A

Public resource group

The public resource group supports only the pay-as-you-go billing method. The following table describes the billing details.

Billing method

Fee calculation

Unit price

Billing duration

Scaling description

Usage notes

Pay-as-you-go

Bill amount = (Price/60) × Usage duration. The usage duration is measured in minutes.

The price of an instance type varies based on the region. We recommend that you view the prices in the PAI console. Perform the following steps: Go to the Create Job page. For more information, see Submit training jobs. In the Resource Configuration section, set the Resource Quota parameter to Public Resources (Pay-As-You-Go). After you select an instance type, you can view the pricing information. image

Running duration of a DLC job

N/A

None

Dedicated resource group

Dedicated resource groups support only the subscription billing method. The following table describes the billing details.

Billing method

Fee calculation

Unit price

Billing duration

Scaling description

Usage notes

Subscription

Bill amount = Prices of instance specifications × Number of instances × Duration.

For more information about the pricing, go to the AI Computing Resource Group (International Site) page.

  • Billing start time: 00:00:00 the next day after the purchase.

  • Billing end time: the time when subscription expires.

N/A

None

Billing examples

Important

The following billing examples are only for reference. Refer to the console or buy page of the service that you want to purchase for the actual fees.

Public resource group

  • Scenario

    For example, you use an instance of the ecs.g6.2xlarge instance type that is deployed in the China (Shanghai) region to create a training job. The job runs for 1 minute and 15 seconds.

  • Fee calculation:

    Bill amount = 1 × 0.6/60 × 1.25 = 0.0125. Unit: USD.

Dedicated resource group

  • Scenario

    For example, you purchase two instances of the ecs.g6.13xlarge-52c192g instance type in the China (Shanghai) region for two months. The subscription unit price is USD 980.57 per month. The price is used as an example.

  • Fee calculation

    Bill amount = 2 × 980.57 × 2 = 3922.28. Unit: USD.

Overdue payments

Causes

The balance of your Alibaba Cloud account is insufficient.

  • Subscription: Your account balance is insufficient to pay for subscription fees.

  • Pay-as-you-go: Your account balance is less than the bill amount of the previous billing cycle. If the system fails to deduct fees, an overdue payment exists in your account.

Service suspension

  • In PAI, bills are generated within 4 hours after the previous billing cycle ends and displayed by using the T + 1 settlement method. If an overdue payment exists in your account, you are notified by text messages.

  • In this case, your PAI resources remain available for 24 hours. After 24 hours, your resources are stopped and your services are suspended.

Note

To ensure service continuity, we recommend that you add funds to your account at the earliest opportunity.

View overdue payments

  1. Log on to the Expenses and Costs console.

  2. On the Account Overview page, view the overdue payments.账户总览

Renewal

Dedicated resource groups support the following renewal methods:

  • Auto-renewal

    To enable auto-renewal, select Auto-renewal when you purchase subscription computing resources. For more information, see Create and manage general computing resources.

  • Manual renewal

    On the dedicated resource group list, find the dedicated resource group that you want to renew and click Renew in the Actions column to renew the resources. For more information, see the "Manage resources" section in the Overview topic.

Refund policies

  • Pay-as-you-go fees cannot be refunded.

  • Subscription fees can be refunded based on the following rules:

    • Five-day money-back guarantee: You are eligible for a full refund for unused resources that are purchased within five days.

    • Partial refund: You are eligible for a refund for the remaining duration. The refund amount is calculated by using the following formula:

      Refund amount = Amount paid - Amount consumed.

    • Refund for renewal orders: You can cancel renewal orders that have not taken effect.

Appendix: Pricing details of the public resource group

The following table provides the pricing details of the instance types supported by DLC. Resources vary based on the regions. The information on the PAI console prevails.

Instance type

Specification

GPU

ecs.g6.xlarge

4 vCPUs + 16 GB memory

None

ecs.c6.large

2 vCPUs + 4 GB memory

None

ecs.g6.large

2 vCPUs + 8 GB memory

None

ecs.g6.2xlarge

8 vCPUs + 32 GB memory

None

ecs.g6.4xlarge

16 vCPUs + 64 GB memory

None

ecs.g6.8xlarge

32 vCPUs + 128 GB memory

None

ecs.r7.large

2 vCPUs + 16 GB of memory

None

ecs.r7.xlarge

4 vCPUs + 32 GB memory

None

ecs.r7.2xlarge

8 vCPUs + 64 GB memory

None

ecs.r7.4xlarge

16 vCPUs + 128 GB memory

None

ecs.r7.6xlarge

24 vCPUs + 192 GB memory

None

ecs.r7.8xlarge

32 vCPUs + 256 GB memory

None

ecs.r7.16xlarge

64 vCPUs + 512 GB memory

None

ecs.g5.xlarge

4 vCPUs + 16 GB memory

None

ecs.g7.xlarge

4 vCPUs + 16 GB memory

None

ecs.g7.2xlarge

8 vCPUs + 32 GB memory

None

ecs.g5.2xlarge

8 vCPUs + 32 GB memory

None

ecs.g6.3xlarge

12 vCPUs + 48 GB memory

None

ecs.g7.3xlarge

12 vCPUs + 48 GB memory

None

ecs.g7.4xlarge

16 vCPUs + 64 GB memory

None

ecs.r7.3xlarge

12 vCPUs + 96 GB memory

None

ecs.c6e.8xlarge

32 vCPUs + 64 GB memory

None

ecs.g6.6xlarge

24 vCPUs + 96 GB memory

None

ecs.g7.6xlarge

24 vCPUs + 96 GB memory

None

ecs.g5.4xlarge

16 vCPUs + 64 GB memory

None

ecs.hfc6.8xlarge

32 vCPUs + 64 GB memory

None

ecs.g7.8xlarge

32 vCPUs + 128 GB memory

None

ecs.hfc6.10xlarge

40 vCPUs + 96 GB memory

None

ecs.g6.13xlarge

52 vCPUs + 192 GB memory

None

ecs.g5.8xlarge

32 vCPUs + 128 GB memory

None

ecs.hfc6.16xlarge

64 vCPUs + 128 GB memory

None

ecs.g7.16xlarge

64 vCPUs + 256 GB memory

None

ecs.hfc6.20xlarge

80 vCPUs + 192 GB memory

None

ecs.g6.26xlarge

104 vCPUs + 384 GB memory

None

ecs.g5.16xlarge

64 vCPUs + 256 GB memory

None

ecs.r5.8xlarge

32 vCPUs + 256 GB memory

None

ecs.re6.13xlarge

52 vCPUs + 768 GB memory

None

ecs.re6.26xlarge

104 vCPU + 1,536 GB memory

None

ecs.re6.52xlarge

208 vCPU + 3,072 GB memory

None

ecs.g7.32xlarge

128 vCPU + 512 GB memory

None

ecs.gn7i-c8g1.2xlarge

8 vCPUs + 30 GB memory

1 × NVIDIA A10

ecs.gn6v-c8g1.2xlarge

8 vCPUs + 32 GB memory

1 × NVIDIA V100

ecs.gn6e-c12g1.24xlarge

96 vCPUs + 736 GB memory

8 × NVIDIA V100

ecs.gn6v-c8g1.16xlarge

64 vCPUs + 256 GB memory

8 × NVIDIA V100

ecs.gn6v-c10g1.20xlarge

82 vCPUs + 336 GB memory

8 × NVIDIA V100

ecs.gn6e-c12g1.12xlarge

48 vCPUs + 338 GB memory

4 × NVIDIA V100

ecs.gn6v-c8g1.8xlarge

32 vCPUs + 128 GB memory

4 × NVIDIA V100

ecs.gn6i-c24g1.24xlarge

96 vCPUs + 372 GB memory

4 × NVIDIA T4

ecs.gn5-c8g1.4xlarge

16 vCPUs + 120 GB memory

2 × NVIDIA P100

ecs.gn7i-c32g1.16xlarge

64 vCPUs + 376 GB memory

2 × NVIDIA A10

ecs.gn6i-c24g1.12xlarge

48 vCPUs + 186 GB memory

2 × NVIDIA T4

ecs.gn6e-c12g1.3xlarge

12 vCPUs + 92 GB memory

1 × NVIDIA V100

ecs.gn5-c4g1.xlarge

4 vCPUs + 30 GB memory

1 × NVIDIA P100

ecs.gn5-c8g1.2xlarge

8 vCPUs + 60 GB memory

1 × NVIDIA P100

ecs.gn5-c28g1.7xlarge

28 vCPUs + 112 GB memory

1 × NVIDIA P100

ecs.gn6i-c4g1.xlarge

4 vCPUs + 15 GB memory

1 × NVIDIA T4

ecs.gn6i-c8g1.2xlarge

8 vCPUs + 31 GB memory

1 × NVIDIA T4

ecs.gn6i-c16g1.4xlarge

16 vCPUs + 62 GB memory

1 × NVIDIA T4

ecs.gn6i-c24g1.6xlarge

24 vCPUs + 93 GB memory

1 × NVIDIA T4

ecs.gn7i-c32g1.8xlarge

32 vCPUs + 188 GB memory

1 × NVIDIA A10

ecs.gn7e-c16g1.4xlarge

16 vCPUs + 125 GB memory

1 × GU50

ecs.gn7-c12g1.3xlarge

12 vCPUs + 95 GB memory

1 × GU50

ecs.gn7i-c16g1.4xlarge

16 vCPUs + 60 GB memory

1 × NVIDIA A10

ecs.gn7-c13g1.26xlarge

104 vCPUs + 760 GB memory

8 × GU50

ecs.ebmgn7e.32xlarge

128 vCPUs + 1,024 GB memory

8 × GU50

ecs.gn7i-c32g1.32xlarge

128 vCPUs + 752 GB memory

4 × NVIDIA A10

ecs.gn7-c13g1.13xlarge

52 vCPUs + 380 GB memory

4 × GU50

ecs.gn7s-c32g1.32xlarge

128 vCPUs + 1,000 GB memory

4 × NVIDIA A30

ecs.gn7s-c56g1.14xlarge

56 vCPUs + 440 GB memory

1 × NVIDIA A30

ecs.gn7s-c48g1.12xlarge

48 vCPUs + 380 GB memory

1 × NVIDIA A30

ecs.gn7s-c16g1.4xlarge

16 vCPUs + 120 GB memory

1 × NVIDIA A30

ecs.gn7s-c8g1.2xlarge

8 vCPUs + 60 GB memory

1 × NVIDIA A30

ecs.gn7s-c32g1.8xlarge

32 vCPUs + 250 GB memory

1 × NVIDIA A30