Dear Alibaba Cloud Customer,

To optimize your experience, reduce bill generation latency, and prevent unexpectedly high outstanding payments during model trials, We plans to roll out a minute-level billing feature for Model Studio model inference calls (excluding batch calls). This update will be launched in phases starting from 10:00 on October 21 (UTC+08), and will enter a public beta period (which will last for some time).

Changes

For billing items related to model inference, excluding batch calls:

1. Reduced deduction latency: The deduction delay (for credit-controlled users, deduction from available balance), which was previously several hours, will be reduced to within approximately 30 minutes. This will help you track your spending more promptly and avoid unexpected outstanding payments.
2. Settlement method adjustment: Model Studio model inference will be settled separately for batch calls and non-batch calls. Additionally, for non-credit-controlled users, a separate deduction will be made each time the accumulated charges for non-batch calls reach $1,000 USD.
3. Billing detail granularity increased to minute-level: Bills will now show minute-level cost details (increased from hour-level), enabling you to analyze and manage your inference costs with greater precision. Detailed billing statements are typically generated within 3-4 hours of usage and will appear within 24 hours at the latest.
4. Downloadable usage data: You can log in to the Alibaba Cloud console and navigate to Billing > Bill Details > View Usage Details to download detailed, request-level usage data for verification.

If you have any questions, contact us by submitting a ticket or calling our service hotline. Thank you for your trust and support in Alibaba Cloud.