This topic describes how to activate AI Guardrails and its pay-as-you-go billing method.
After you activate the pay-as-you-go plan for AI Guardrails, the Content Compliance check service is enabled by default. To check for sensitive content or prompt attacks, you must enable these features in the console. For more information, see the instructions in this topic.
Activate the pay-as-you-go plan for AI Guardrails
Activating the pay-as-you-go plan for AI Guardrails is free of charge. You are billed based on your actual usage. The following table describes the pricing details.
Connection type | Moderation type | Supported business scenarios (services) | Unit price |
API/AI Gateway integration | Text moderation - Advanced (per call) text_guard_advanced |
| USD 6 per 10,000 calls |
API/AI Gateway integration | Sensitive content moderation - Advanced (per call) text_sddp_advanced |
| USD 6 per 10,000 calls |
API/AI Gateway integration | Prompt attack moderation |
| This feature is in public preview. After you enable it, you can use it for free. |
The pay-as-you-go plan for AI Guardrails bills you based on the type of content checked and the usage volume.
Go to the AI Guardrails product activation page.
Read and select the Terms of Service, and then click Activate Now.
Overdue payments and top-ups
The billing cycle for AI Guardrails is 24 hours.
Alibaba Cloud generates a bill for your usage from the previous billing cycle and deducts the amount from your Alibaba Cloud account. Bills are typically generated 10 to 12 hours after the end of each calendar day.
If your account has an overdue payment, the service is suspended. After you pay the overdue amount, the service resumes within 5 minutes.
Enable sensitive content detection, prompt attack detection, malicious file detection, or digital watermarking
Log on to the AI Guardrails console. On the page, enable the detection features that you need.
If you want to enable Sensitive content detection, read the message in the dialog box. This feature is billed separately.

If you want to enable Prompt Injection Attack detection, read the message in the dialog box. This feature is in public preview and available as a free trial. If you have any questions, contact your business manager.
