AI Guardrails:Multi-language services in Text Moderation 2.0

Level-1 labels

Subcategory labels

Subcategory labels (riskTips) provide finer-grained classification in xxx_yyy format. For example, contraband_Drugs indicates that the contraband content relates specifically to drugs.

Risk levels

The riskLevel field indicates the severity of detected content and the recommended action.

Content with no detected risk returns an empty Labels field — handle it based on your business requirements.

Step 1: Activate the service

Open the service activation2.0 page to activate Text Moderation 2.0. After activation, the default billing method is pay-as-you-go.

Step 2: Grant permissions to a RAM user

Before calling API operations as a Resource Access Management (RAM) user, grant the user the required permissions.

1. Log on to the RAM console as a RAM administrator.

2. Create a RAM user. For details, see Create a RAM user.

3. Grant the AliyunYundunGreenWebFullAccess system policy to the RAM user. For details, see Grant permissions to a RAM user.

After completing these steps, the RAM user can call the Content Moderation API. To authenticate API calls, create an AccessKey pair for the RAM user and include it in each request. For details, see Obtain an AccessKey pair.

Step 3: Install an SDK and call the API

The following table lists the supported regions and their endpoints. For SDK installation instructions and usage examples, see Text Moderation 2.0 SDKs and usage guide.

To get SDK sample code for other programming languages, use OpenAPI Explorer. OpenAPI Explorer dynamically generates sample code for each supported SDK.

Usage notes

Service endpoint: https://green-cip.{region}.aliyuncs.com

Call the TextModeration operation to submit a text moderation task. For details on constructing HTTP requests, see Request syntax.

Billing: Only requests that return HTTP status code 200 are charged.

QPS limit: 100 calls per second per account. Requests that exceed this limit trigger throttling.

Request parameters

ServiceParameters fields

Response parameters

Data object fields

Reason JSON fields

Examples

All examples use the comment_multilingual_global service. Replace accountId with your actual account ID.

Request

{
    "Service": "comment_multilingual_global",
    "ServiceParameters": {
        "content": "testing content",
        "accountId": "10123****",
        "dataId": "text0424****"
    }
}

Response: custom library and risky words both detected

{
    "Code": 200,
    "Data": {
        "Labels": "C_customized,profanity",
        "Reason": "{\"riskLevel\":\"high\",\"customizedWords\":\"HXXXXX\",\"detectedLanguage\":\"en\",\"riskTips\":\"sexuality_Suggestive\",\"riskWords\":\"pxxxxy\",\"translatedContent\":\"pxxxxy sxxxx,HXXXXX\",\"customizedLibs\":\"ZXXXXX\"}"
    },
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Response: custom library only

{
    "Code": 200,
    "Data": {
        "Labels": "C_customized",
        "Reason": "{\"riskLevel\":\"high\",\"customizedWords\":\"HXXXXX\",\"detectedLanguage\":\"en\",\"translatedContent\":\"HXXXXX\",\"customizedLibs\":\"ZXXXXX\"}"
    },
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Response: risky words only

{
    "Code": 200,
    "Data": {
        "Labels": "sexuality",
        "Reason": "{\"riskLevel\":\"high\",\"detectedLanguage\":\"ar\",\"riskTips\":\"sexuality_Suggestive\",\"riskWords\":\"pxxxxy\",\"translatedContent\":\"pxxxxy sxxxx\"}"
    },
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Response: no risk detected

{
    "Code": 200,
    "Data": {
        "Labels": "",
        "Reason": "{\"detectedLanguage\":\"en\",\"translatedContent\":\"AXXXXXX\"}"
    },
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Error codes