June 2025 promotion: Model Studio LLM packages

No. For details, see Usage and purchase limits.

Different resource packages apply to different models:

• The qwen-max Large Language Model Inference resource package can only be used to pay for the real-time inference fees of the qwen-max and qwen-max-latest models (non-thought mode, including input and output). This means that in model calls, the value of the model parameter is qwen-max or qwen-max-latest.
• The qwen-plus Large Language Model Inference resource package can only be used to offset the real-time inference fees for the qwen-plus and qwen-plus-latest models when a single request has an input token count in the 0<Token≤128K range (for non-thought mode, including both input and output tokens). This means that in model calls, the value of the model parameter is qwen-plus or qwen-plus-latest.
• The Qwen-Turbo Large Language Model Inference resource package can be used to offset fees only for the real-time inference (non-thought mode, including input and output) of the qwen-max and qwen-turbo-latest models. This means that when you call a model, the model parameter must be set to qwen-max or qwen-turbo-latest.

None of these resource packages cover snapshot version models.

Free quota is always used before resource package quota.

The Qwen-plus resource package from March only covers the qwen-plus model, so the deduction order is as follows:

• For real-time inference costs from the qwen-plus model, your March Qwen-plus resource package will be used first. Once it is depleted, your newly purchased Qwen-plus resource package will be used.
• For real-time inference costs from the qwen-plus-latest model, your newly purchased resource package will be used.

The Qwen-max resource package purchased in March covers both qwen-max and qwen-max-latest models. Therefore:

• For real-time inference costs from either the qwen-max or qwen-max-latest models, your March Qwen-max resource package will be used first. Once it is depleted, your newly purchased resource package will be used.

Follow these steps to confirm which resource package was used:

1. On the bill details page, set Cycle to Details. Then select a Billing Month, set Product Name to Alibaba Cloud Model Studio (Bailian), and click Search.

2. Sort the results by Billing Date in descending order (latest first), and then view and record the Deducted Resource Pack ID.

   In the bill details, the Resource Pack Deduction column shows the amount deducted for each billing item (for example, 0.503 or 0.272), which confirms that the resource package was applied. The Deducted Resource Pack ID column (for example, sfm_llminference_dp...) identifies the specific resource package used. Record this ID for verification.

3. In Orders, set Product Name to Alibaba Cloud Model Studio (Bailian). Then enter the ID you recorded in the previous step in Asset/Resource Pack Instance ID and click Search.

   In the results, check the Payment/Activation Time column to confirm the purchase date of each resource package.

4. In the search results, view the Payment/Activation Time of the resource package.

   Resource packages purchased in March: Orders placed between 18:00:00, February 28, 2025, and 00:00:00, April 1, 2025.

   Resource packages purchased during this promotion: Orders placed between 17:00:00, May 30, 2025, and 00:00:00, July 1, 2025.

On the bill details page, set Cycle to Details. Then select a Billing Month, set Product Name to Alibaba Cloud Model Studio (Bailian), and click Search.

Sort the results by Billing Date in descending order (latest first). The Asset/Resource Instance ID column identifies which model was used, and the Resource Pack Deduction column shows the number of tokens deducted.

The Asset/Resource Instance ID format is semicolon-delimited and includes information such as the model name, token type, and call method. For example, qwen-plus;input_token;app indicates the model is qwen-plus, the token type is input_token, and the call method is app.

The resource package is valid for one year starting from the date of purchase, not the date it is first used.

If you have a previously purchased, unexhausted resource package of the same type, it will be used first. To avoid your new resource package expiring, we recommend using your quota promptly.

For example, you purchase Qwen-Max resource package A on March 5, which has quota remaining. On June 5, you purchase Qwen-Max resource package B. Deductions will continue to come from package A until it is depleted, so deductions for package B might not begin until August 30. However, package B is valid for one year from its purchase date (June 5), not its first-use date (August 30).

This error may indicate that your Alibaba Cloud account has been suspended due to an overdue payment. Even if you have a valid resource package, you cannot make API calls if your account has an outstanding balance. Please go to the Billing page to check for overdue payments and add funds to your account.

Your inference request may have triggered the context cache feature. Resource packages do not cover costs associated with context cache. You can check your bill details to see if the instance ID associated with the overdue payment contains "input_token_cache".

No. A resource package can only cover usage that occurs after its purchase. You must add funds to your account to pay any overdue balance.

Yes. You can initiate a refund request on the Unsubscribe page. Alibaba Cloud will issue a refund based on the unsubscription rules.

1. On the Unsubscribe page, on the Resource Plan tab, select the appropriate "Large language model inference resource package xxx" for Product Name. For Type, select either Unsubscription for Unused Resource Packs Within Five Days or Non-full Amount Unsubscription, then click Search.

   • Unsubscription for Unused Resource Packs Within Five Days: Select this if it has been 5 days or less since purchase and the package is completely unused.
   • Non-full Amount Unsubscription: Select this if it has been more than 5 days since purchase or if the package has been partially used.

2. In the search results, find the resource package instance you want to refund and click Unsubscribe in its row. Alternatively, you can select multiple instances and click Batch Unsubscribe.

3. On the Unsubscribe page, carefully review the resource information, terms, and refund amount. Follow the on-screen instructions to complete the process.

4. After the unsubscription is complete, you can find the order in the Unsubscription Records.

5. Click Details in the Actions column of the order to view information such as Refund Status, Fund Flow, Refund Time, and Refund Amount.

To check your remaining quota, log in to the Alibaba Cloud console and navigate to the Billing page. From the main Billing page, click Resource Packs in the left-side navigation pane. In the resource dimension, select Resource Packs and filter the status to Valid. You can then view the remaining quota in the Total Amount and Remaining Amount columns.

No notifications are sent when a resource package is depleted. We recommend monitoring your usage in the Alibaba Cloud bill details.

Log in to the Alibaba Cloud Model Studio console and click Enable Now in the upper-right corner to activate Model Studio and receive your free quota. Your free quota will be used first. After it is exhausted, your resource package will be used for deductions.

This is a limited-quantity offer. A total of 2,000 Qwen-Max packages, 3,000 Qwen-Plus packages, and 3,000 Qwen-Turbo packages are available. The offer is valid while supplies last.