Documentation
Server APIs v2
Streaming moderation
Start moderating audio streams

Start moderating audio streams

Last updated：2025-07-10 10:38

1 Description

Audio stream moderation is an interactive extension service of ZEGOCLOUD's real-time audio and video products. Through the audio stream moderation interface, developers can call real-time audio moderation, including audio semantic recognition and audio feature (pitch, timbre, voiceprint, melody, etc.) recognition. ZEGOCLOUD will send the recognition results to developers through Callback on moderated audio streams or configured custom callback addresses. The Start moderating audio streams feature should be used in conjunction with Stop moderating audio streams.

Before using this method for the first time, you need to contact ZEGOCLOUD technical support to activate it.

2 Request method and endpoint

Request method: POST
When using POST request method to pass parameters:
- Parameters in the Body can be passed directly in JsonObject format without serializing to String format.
- In Headers, set "Content-type" to "application/json".

Request endpoint: https://rtc-api.zego.im/?Action=StartCensorAudioV2

Protocol: HTTPS

Rate limit (all rooms with the same AppID): 100 requests/second

3 Request parameters

Listed below are the parameters specific to this request and part of the public request parameters. For the complete list of public request parameters, see Server APIs public request parameters.

Parameter	Type	Required	Description
RoomId	String	Yes	The room ID that needs audio moderation. By default, all streams in the room will be submitted for moderation. If there are streams in the room that do not need moderation (such as streams from trusted users), please set `ZegoPulisherConfig.streamCensorshipFlag` to disallow moderation when calling the client SDK `startPublishingStream`.
EventId	String	Yes	Scene identifier. This parameter identifies the scene corresponding to the audio stream. The moderation strategy can be adjusted according to the scene to achieve control of different scene moderation standards. Live Broadcasting: live_streaming Audio Chat Room: audio_room 1v1: private_chat If the above optional values cannot fully meet your usage requirements, please contact ZEGOCLOUD technical support.
ResultCallbackUrl	String	No	Custom callback address for moderation results. If you need to send the moderation results of this moderation task through another callback address, you can achieve this through this parameter. If not needed, please contact ZEGOCLOUD technical support for processing.
StatusCallbackUrl	String	No	Custom callback address for moderation status. If you need to send the moderation status of this moderation task through another callback address, you can achieve this through this parameter. If not needed, please contact ZEGOCLOUD technical support for processing.
IsMixingEnabled	Int32	No	Whether to submit for moderation through stream mixing. 0: Default value, no stream mixing moderation, i.e., single stream moderation. Each audio stream in the room is moderated independently. 1: Stream mixing moderation, all audio streams in the room are mixed into one stream. Stream mixing moderation can save audio moderation costs, but correspondingly makes it more difficult to accurately identify violating streams and users. Please choose the appropriate moderation method based on your business needs.
AudioLanguage	String	Yes	Audio stream language. When semantic recognition of audio streams is needed, please input this value accurately. If the following languages cannot meet your needs, please contact technical support for configuration. zh: Chinese, ja: Japanese, ko: Korean, th: Thai vi: Vietnamese, ms: Malay, tl: Filipino, id: Indonesian hi: Hindi, ar: Arabic, tr: Turkish, en: English es: Spanish, pt: Portuguese, it: Italian, fr: French de: German, ru: Russian
RiskTypeList	Array of String	No	Risk type list. At least one of RiskTypeList and BusinessTypeList must not be empty. The following values can be added to the array: Audio Semantics: EROTIC: Pornography DIRTY: Abuse POLITY: Political BAN: Prohibited VIOLENT: Violence and terrorism ADVERT: Advertisement Audio Features: MOAN: Moan AUDIOPOLITICAL: Leader voiceprint recognition ANTHEN: National anthem recognition BANDAUDIO: Prohibited songs
BusinessTypeList	Array of String	No	Business recognition categories. At least one of RiskTypeList and BusinessTypeList must not be empty. The following values can be added to the array: Audio Semantics: MOINIR: Minor Audio Features: SING: Singing LANGUAGE: Language recognition VOICE: Voice attribute recognition Audio stream recognition is not limited to violation content, it can also assist developers in business operations. If the above optional values cannot meet your business needs, please contact ZEGOCLOUD technical support.
LabelLanguage	String	Yes	Language type of labels in callback information (Label related). zh: Chinese en: English
ReturnAllText	Int32	No	Whether to return all text for the current 10s segment. The risk levels of moderation results will be classified into three categories: PASS: Normal content, recommended to pass directly. REVIEW: Suspicious content, recommended for manual review. REJECT: Violating content, recommended to block directly. 0: Return audio segment text with non-pass risk level (default value). 1: Return audio text segments of all risk levels.
ReturnPreAudio	Int32	No	Whether to return the audio segment link before the violating segment. 0: Do not return the audio segment link before the violating segment, only return the violating segment audio link. (Default value) 1: Return the audio segment link before the violating segment.
ReturnFinishInfo	Int32	No	Whether to perform "moderation task status" callback when the moderation task ends. 0: Do not send end notification when moderation ends. 1: Initiate end notification when moderation ends.

4 Sample request

Request URL

https://rtc-api.zego.im/?Action=StartCensorAudioV2
&<Common request parameters>

Request body

{
    "RoomId": "room_1",
    "IsMixingEnabled": 0,
    "EventId": "live_streaming",
    "RiskTypeList": [
        "ADLAW",
        "ADVERT",
        "AUDIOPOLITICAL",
        "BAN",
        "DIRTY",
        "EROTIC",
        "MOAN",
        "POLITY",
        "VIOLENT"
    ],
    "AudioLanguage": "zh",
    "LabelLanguage": "zh",
    "ReturnAllText": 1,
    "ReturnPreText": 1,
    "ReturnPreAudio": 1,
    "ReturnFinishInfo": 1,
    "ResultCallbackUrl": "",
    "StatusCallbackUrl": ""
}

5 Response parameters

Parameter	Type	Description
Code	Int32	Return code. 50006, 50007: HTTP request failed, please try again later (recommended 300s). If it still fails, please contact ZEGOCLOUD technical support. 50009: Moderation failed, please try again later (recommended 300s), or contact ZEGOCLOUD technical support. 50117: Configuration error, Shumei audio stream moderation permission not enabled, please contact ZEGOCLOUD technical support.
Message	String	Operation result description.
RequestId	String	Request ID.
Data	Object	Response data.
TaskId	String	Moderation task ID, used to end the moderation task.

6 Sample response

{
    "Code":0,
    "Message":"success",
    "RequestId":"TestRequestId1635941437248027000",
    "Data":{
              "TaskId":"15515d8fef61801b05c22e0ebbe8681f"
           }
}

7 Return codes

Listed below are the return codes related to this API. For the complete list of return codes, see Return codes.

Return Code	Description	Handling Suggestion
50006, 50007	HTTP request failed.	Please try again later (recommended 300s), or contact ZEGOCLOUD technical support.
50009	Moderation failed.	Please try again later (recommended 300s), or contact ZEGOCLOUD technical support.
50117	Configuration error, Shumei permission not enabled.	Please contact ZEGOCLOUD technical support.