AI Guardrails:LLM Voice Moderation

Overview

Voice Moderation (Large Model Edition) uses the Qwen LLM, fine-tuned for audio content risk detection. The following services are available:

• Audio and Video Media Detection_LLM Edition:

  Moderates audio files. Uses the Qwen large model for enhanced speech recognition and risk detection. Supports 26 languages, including Chinese, English, Japanese, German, Korean, Russian, French, Portuguese, Arabic, Italian, Spanish, Hindi, Indonesian, Thai, Turkish, Ukrainian, Vietnamese, Czech, Danish, Filipino, Finnish, Icelandic, Malay, Norwegian, Polish, and Swedish.

• Social & Entertainment Live Stream Detection_LLM Version:

  Moderates live audio streams. Uses the Qwen large model for enhanced speech recognition and risk detection. Supports 26 languages, including Chinese, English, Japanese, German, Korean, Russian, French, Portuguese, Arabic, Italian, Spanish, Hindi, Indonesian, Thai, Turkish, Ukrainian, Vietnamese, Czech, Danish, Filipino, Finnish, Icelandic, Malay, Norwegian, Polish, and Swedish.

Service selection

Billing

Voice Moderation (Large Model Edition) supports pay-as-you-go billing and resource packages.

Pay-as-you-go

Pay-as-you-go is the default billing method. You are billed based on actual usage, with no charges when the service is idle.

Get started

Step 1: Activate the service

Go to the Voice ModerationVoice Moderation activation page to activate the service.

Pay-as-you-go billing applies by default — no charges when idle. Purchase a resource package for tiered discounts on high-volume usage.

Step 2: Grant permissions to a RAM user

Before calling the API, grant permissions to a RAM user. API calls require an AccessKey pair from your Alibaba Cloud account or a RAM user. For more information, see Obtain an AccessKey.

1. Log on to the RAM console using your Alibaba Cloud account or a RAM administrator account.

2. Create a RAM user. For more information, see Create a RAM user.

3. Attach the AliyunYundunGreenWebFullAccess system policy to the RAM user. For more information, see Grant permissions to a RAM user. The RAM user can then call Content Moderation API operations.

Step 3: Install and integrate the SDK

Follow the Voice Moderation SDK and Integration Guide to install and integrate the SDK. The service is available in the following regions:

Step 4: Adjust moderation rules (Optional)

In the Content Moderation console, configure detection rules, manage text moderation rules, clone services, set up custom dictionaries, view moderation logs, and monitor usage.

API reference

Usage notes

Submit audio content moderation tasks with this API. To construct HTTP requests, see Making native HTTP calls. For SDK usage, see the Voice Moderation SDK and integration guide.

• APIs:

  • VoiceModeration: Submit an Voice Moderation task.

  • VoiceModerationResult: Query moderation task results.

  • VoiceModerationCancel: Cancel a moderation task.

• Billing:

  Paid API. Only requests returning HTTP status code 200 are billed. For more information, see Billing.

• Service performance:

  Service performance	Description
  Audio file size	Supports audio files up to 500 MB.
  Audio and video file formats	Supported audio file formats: MP3, WAV, AAC, WMA, OGG, M4A, and AMR. Supported video file formats: AVI, FLV, MP4, MPG, ASF, WMV, MOV, RMVB, and RM.
  Live audio stream	Supported protocols: RTMP, HLS, HTTP-FLV, and RTSP.
  Queries per second (QPS)	100 QPS for task submission.
  Concurrent streams	50 concurrent streams by default.

Submit a moderation task

Request parameters

Response parameters

Examples

• Request example

{
    "Service": "audio_detection_byllm_global",
    "ServiceParameters": {
        "cryptType": "SHA256",
        "seed": "abc***123",
        "callback": "https://aliyun.com/callback",
        "url": "http://aliyundoc.com/test.mp3"
    }
}

• Successful response example

{
    "Code": 200,
    "Data": {
        "TaskId": "AAAAA-BBBBB",
        "DataId": "voice20240307***"
    },
    "Message": "SUCCESS",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Query task results

For an ongoing live stream moderation task, a query returns the most recent audio slices. After the task completes, the query returns all slices.

• To query a moderation task, call the VoiceModerationResult operation.

• Billing: This API is free of charge.

• Query timing: Wait at least 30 seconds after submitting an asynchronous task before querying results. Results are retained for 24 hours and then automatically deleted.

Response parameters

Examples

Request example

{
    "Service": "audio_detection_byllm",
    "ServiceParameters": {
        "taskId": "AAAAA-BBBBB"
    }
}

Successful response example

{
    "Code": 200,
    "Data": {
        "DataId": "voice20240307***",
        "LiveId": "liveId1****",
        "RiskLevel": "high",
        "SliceDetails": [
            {
                "StartTime": 0,
                "EndTime": 4065,
                "RiskLevel": "high",
                "Result": [
                    {
                        "Label": "political_entity",
                        "Description": "Suspected political entity",
                        "Confidence": 100.0,
                        "RiskLevel": "high",
                        "RiskWords": "WordA,WordB",
                        "RiskPositions": [
                            {
                                "EndPos": 14,
                                "RiskWord": "WordA",
                                "StartPos": 16
                            }
                        ]
                    }
                ],
                "Text": "Content Moderation product test case",
                "Url": "https://aliyundoc.com"
            }
        ]
    },
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Cancel a moderation task

You can only cancel live stream moderation tasks. File-based moderation tasks cannot be canceled.

• To cancel a moderation task, call the VoiceModerationCancel operation.

• Billing: This API is free of charge.

Request parameters

Response parameters

Examples

• Request example

{
    "Service": "live_detection_byllm_global",
    "ServiceParameters": {
        "taskId": "AAAAA-BBBBB"
    }
}

• Successful response example

{
    "Code": 200,
    "Message": "OK",
    "RequestId": "AAAAAA-BBBB-CCCCC-DDDD-EEEEEEEE****"
}

Callback message format

Status codes

The following table lists the status codes returned by the API. You are billed only for successful requests (code 200).