Documentation
Server APIs v2
Server callbacks
Callback on moderated audio streams

Callback on moderated audio streams

Last updated：2025-05-22 16:15

1 Description

When developers need audio stream content moderation, after starting an Start moderating audio streams task, ZEGOCLOUD will send POST requests to the callback address to provide audio stream moderation results. When the ReturnFinishInfo parameter in the Start moderating audio streams task is set to 1, a moderation task status callback will be initiated after the moderation task ends.

2 Callback Description

Request method: POST

The callback data format is JSON, and you need to perform UrlDecode decoding on it.

Request URL: Please contact ZEGOCLOUD technical support to configure the callback URL.
Protocol: HTTPS/HTTP, HTTPS recommended.

3 Callback Parameters

3.1 Moderation Result Callback

Common Parameters	Type	Description
Event	String	Callback event, returns `censor_audio_v2_result` for this callback.
AppId	Number	AppID.
Timestamp	Number	Server current time, Unix timestamp (seconds).
Nonce	String	Random number.
Signature	String	Signature.
Business Parameters	Type	Description
Code	Number	Return code. 0 indicates success.
Message	String	Operation result description.
TaskId	String	Moderation task ID. Corresponds to the TaskId returned in Start moderating audio streams response parameters.
ResultTaskId	String	ID of this callback information.
Detail	Object	Moderation result details for the 10s segment.
RiskLevel	String	Risk level (exists when code is 0). Developers can handle the violating segment's audio stream and users based on this value. Note: This field will return the risk level of the highest risk "Risk Type List" in the 10s audio (e.g., "Political" usually has a higher risk level). The risk level priority of specific "Risk Type Lists" is determined by Shumei's internal moderation system. PASS: Normal content, recommended to pass directly. REVIEW: Suspicious content, recommended for manual review. REJECT: Violating content, recommended to block directly.
AudioText	String	Semantic text corresponding to the 10s audio segment.
AudioUrl	String	URL of the 10s audio segment.
PreAudioUrl	String	Only has value when ReturnPreAudio=1 in the start moderation task request, represents the URL of a 20s audio segment containing the current violating audio segment + the previous audio segment.
RiskDescription	String	Specific "Risk Type List" and subcategories of the highest risk violation content in the current 10s segment. Will be categorized by detail level into Level 1 tags (e.g., Political), Level 2 tags (e.g., Top Leader), and Level 3 tags (e.g., Top Leader Semantics). Note: Only for reference when understanding risk reasons, please do not rely on this parameter's value for logical processing. Returns "Normal" when RiskLevel is PASS. Returns custom list name when hitting custom lists. Other cases format: Level1Tag:Level2Tag:Level3Tag.
RiskLabel1	String	Level 1 tag. Returns "normal" when RiskLevel is PASS.
RiskLabel2	String	Level 2 tag. Level 2 tags belong to Level 1 tags, empty when riskLevel is PASS.
RiskLabel3	String	Level 3 tag. Level 3 tags belong to Level 2 tags, empty when riskLevel is PASS.
VadStatus	Number	Audio segment silence status: 0: Silent segment. When silent, other field values will be empty, subject to actual callback content. 1: Non-silent segment.
RiskDetail	Object	Risk details. Will show the highest risk information in the 10s segment. For example, if the current 10s segment hits both "Pornography" and "Political" "Risk Type Lists", and the backend determines "Political" has higher risk than "Pornography", then "Political" related risk details will be returned.
RiskSource	Number	Risk source: 1000: No risk. 1001: Audio semantics. 1003: Audio features (e.g., pitch, timbre, voiceprint, melody, etc.).
AudioText	String	Text transcribed from 10s audio semantics.
MatchedLists	Array of Object	Custom lists hit (configured by contacting technical support).
Name	String	Custom list name.
RiskInfoList	Array of Object	List of all risk information in the 10s audio segment. Sorted by Shumei's internally defined risk level priority from high to low.
RiskLevel	Number	Risk level (exists when code is 0). Developers can handle the violating segment's audio stream and users based on this value. PASS: Normal content, recommended to pass directly. REVIEW: Suspicious content, recommended for manual review. REJECT: Violating content, recommended to block directly.
RiskDescription	String	Specific "Risk Type List" and subcategories of the highest risk violation content in the current 10s segment. Will be categorized by detail level into Level 1 tags (e.g., Political), Level 2 tags (e.g., Top Leader), and Level 3 tags (e.g., Top Leader Semantics). Note: Only for reference when understanding risk reasons, please do not rely on this parameter's value for logical processing. Returns "Normal" when RiskLevel is PASS. Returns custom list name when hitting custom lists. Other cases format: Level1Tag:Level2Tag:Level3Tag.
RiskLabel1	String	Level 1 tag. Returns "normal" when RiskLevel is PASS.
RiskLabel2	String	Level 2 tag. Level 2 tags belong to Level 1 tags, empty when riskLevel is PASS.
RiskLabel3	String	Level 3 tag. Level 3 tags belong to Level 2 tags, empty when riskLevel is PASS.
RiskDetail	Object	Risk details, will show the highest risk information in the 10s segment. For example, if the current 10s segment hits both "Pornography" and "Political" "Risk Type Lists", and the backend determines "Political" has higher risk than "Pornography", then "Political" related risk details will be returned.
RiskSource	Number	Risk source: 1000: No risk. 1001: Audio semantics. 1003: Audio features (e.g., pitch, timbre, voiceprint, melody, etc.).
AudioText	String	Text transcribed from 10s audio semantics. Only has value for the highest risk information.
MatchedLists	Array of Object	Custom lists hit (configured by contacting technical support).
Name	String	Custom list name.
BusinessInfoList	Array of Object	List of all business information. If the following fields cannot meet your needs, please contact technical support for adjustments.
BusinessDescription	String	Business tag description in Chinese. Format is "Level1BusinessTag:Level2BusinessTag:Level3BusinessTag" in Chinese, e.g., Portrait:Portrait Pose:Sitting.
BusinessLabel1	String	Level 1 business tag.
BusinessLabel2	String	Level 2 business tag.
BusinessLabe3	String	Level 3 business tag.
AuxInfo	Object	Auxiliary information.
RoomId	String	Room ID for moderation.
ProcessBeginTime	Number	Time when 10s audio moderation started (13-digit Unix timestamp).
ProcessFinishTime	Number	Time when 10s audio moderation ended (13-digit Unix timestamp).

3.2 Moderation Task Status Callback

When the ReturnFinishInfo parameter in the Start Audio Stream Moderation task is set to 1, a moderation task status callback will be initiated after the audio stream moderation task ends.

Common Parameters	Type	Description
Event	String	Callback event, returns `censor_audio_v2_status` for this callback.
AppId	Number	AppID.
Timestamp	Number	Server current time, Unix timestamp (seconds).
Nonce	String	Random number.
Signature	String	Signature.
Business Parameters	Type	Description
Code	Number	Return code, 0 indicates success.
Message	String	Operation result description.
TaskId	String	Moderation task ID, corresponds to the TaskId returned by Start moderating audio streams.
Status	Number	Moderation status. 0: Moderation ended.
AuxInfo	Object	Auxiliary information.
RoomId	String	Room ID.
CensorStreamTime	Number	Total duration of the stream moderated in this task (unit: seconds).

4 Sample Data

4.1 Moderation Result Callback Example

{
    //Common parameters
    "Event": "censor_audio_v2_result",
    "AppId": 1,
    "Timestamp": 1724743250,
    "Nonce": "7407715855877898783",
    "Signature": "5cc9e67af0ba0c95f99bd73f79a36485f574ad11"
    //Business parameters
    "Code": 0,
    "Message": "success",
    "TaskId": "384a8a77aeb352d3ec8144ab4640cc52",
    "ResultTaskId": "384a8a77aeb352d3ec8144ab4640cc52_2",
    "Detail": {
        "RiskLevel": "REJECT",
        "AudioText": "Example of violating content",
        "AudioUrl": "http://xxxx/POST_AUDIOSTREAM%2FMP3%2F20240606%2F384a8a77aeb352d3ec8144ab4640cc52_2.mp3?Expires=1720269317&OSSAccessKeyId=LTAI5tLsVBxJ8nhyy5gQVW3K&Signature=8JZymbV%2F6Psm72k6S2Xq3Dcrg14%3D",
        "PreAudioUrl": "http://xxxx/POST_AUDIOSTREAM%2FMP3%2F20240606%2F384a8a77aeb352d3ec8144ab4640cc52_2_pre.mp3?Expires=1720269317&OSSAccessKeyId=LTAI5tLsVBxJ8nhyy5gQVW3K&Signature=BKUDNNWPklQldaEMSFOvSts6O84%3D",
        "RiskDescription": "Advertisment",
        "RiskLabel1": "ad",
        "RiskLabel2": "ad",
        "RiskLabel3": "ad",
        "VadStatus": 1,
        "RiskDetail": {
            "RiskSource": 1001,
            "AudioText": "Example of violating content",
            "MatchedLists": null
        }
    },
    "AuxInfo": {
        "RoomId": "room_1",
        "ProcessBeginTime": 1717677317155,
        "ProcessFinishTime": 1717677317554
    }
}

4.2 Moderation Task Status Callback Example

{
    //Common parameters
    "Event": "censor_audio_v2_status",
    "AppId": 1,
    "Timestamp": 1724743250,
    "Nonce": "7407715855877898783",
    "Signature": "5cc9e67af0ba0c95f99bd73f79a36485f574ad11"
    //Business parameters
    "Code": 0,
    "Message": "success",
    "TaskId": "384a8a77aeb352d3ec8144ab4640cc52",
    "Status": 0,
    "AuxInfo": {
        "RoomId": "room_1",
        "CensorStreamTime": 31
    }
}

5 Response

A returned HTTP status code of 2XX (e.g., 200) indicates success, all other responses indicate failure.

6 Callback Retry Strategy

If the ZEGOCLOUD server does not receive a response, or the HTTP status code received by the ZEGOCLOUD server is not 2xx (such as 200), it will retry with intervals of 2s, 4s, 8s, 16s, and 32s respectively. If the retry still fails after an interval of 32s, it will no longer retry.