tencent cloud

Mobile Live Video Broadcasting

ドキュメントMobile Live Video BroadcastingClient APIsiOSAITranscriberStore

AITranscriberStore

Download
フォーカスモード
フォントサイズ
最終更新日: 2026-07-16 14:59:12

Introduction

AITranscriberStore provides a complete set of AI real-time transcription management APIs, including starting, updating, and stopping real-time transcription. Through this class, you can implement real-time speech-to-text transcription with multi-language translation support during audio/video calls or live streaming.
Important:
Use the create(roomID:) factory method to create an AITranscriberStore instance for a specific room. Configure the source language and translation languages before starting transcription.
Note:
Transcription state updates are delivered through the state publisher. Subscribe to it to receive real-time transcription messages.

Features

Real-time Transcription: Supports real-time speech-to-text transcription during audio/video sessions.
Multi-language Source: Supports 20 source languages including Chinese, English, Japanese, Korean, etc.
Real-time Translation: Supports translating transcribed text into 15 target languages in real-time.
State Management: Provides real-time message list state for UI display.

API List

Function
Description
Create an instance for a specific room.
Start transcription.
Update transcription configuration.
Stop transcription.
Start interpretation.
Update interpretation configuration.
Stop interpretation.
Set interpretation volume.

Getting Instance

create(roomID:)

Create an instance for a specific room.

Transcription Control

startTranscription

Start transcription.
public func startTranscription(myLanguage: SourceLanguage, config: TranscriptionConfig = TranscriptionConfig(), completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Start the AI transcription service with the specified language and configuration. Once started, the transcriber will recognize speech in the room and provide real-time transcription results through the state publisher.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
myLanguage
The source language for the current user's speech recognition.
config
Transcription configuration, including whether to enable translation.
completion
CompletionClosure?
Completion callback (success/failure status).

updateTranscription

Update transcription configuration.
public func updateTranscription(myLanguage: SourceLanguage, config: TranscriptionConfig = TranscriptionConfig(), completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Update the configuration of the ongoing transcription service. This can be used to change the source language or modify the translation settings without stopping and restarting the transcription.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
myLanguage
The new source language for speech recognition.
config
New transcription configuration.
completion
CompletionClosure?
Completion callback (success/failure status).

stopTranscription

Stop transcription.
public func stopTranscription(completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Stop the ongoing AI transcription service. After stopping, the transcriber will no longer recognize speech or provide transcription results.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
completion
CompletionClosure?
Completion callback (success/failure status).

Interpretation Control

startInterpretation

Start interpretation.
public func startInterpretation(myLanguage: SourceLanguage, config: InterpretationConfig = InterpretationConfig(), completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Start the AI interpretation service with the specified language and configuration. Once started, the interpreter will provide real-time voice interpretation in the room.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
myLanguage
The source language for interpretation.
config
Interpretation configuration, including voice type.
completion
CompletionClosure?
Completion callback (success/failure status).

updateInterpretation

Update interpretation configuration.
public func updateInterpretation(myLanguage: SourceLanguage, config: InterpretationConfig = InterpretationConfig(), completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Update the configuration of the ongoing interpretation service.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
myLanguage
The new source language for interpretation.
config
New interpretation configuration.
completion
CompletionClosure?
Completion callback (success/failure status).

stopInterpretation

Stop interpretation.
public func stopInterpretation(completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Stop the ongoing AI interpretation service.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
completion
CompletionClosure?
Completion callback (success/failure status).

setInterpretationVolume

Set interpretation volume.
public func setInterpretationVolume(_ volume: Int, completion: CompletionClosure? = nil) {
    fatalError("\\(#function) must be overridden by subclass")
}
Set the volume level for the interpretation audio output.
Version
Supported since version 3.5.
Parameters
Parameter
Type
Description
volume
Int
Volume level.
completion
CompletionClosure?
Completion callback (success/failure status).

Data Structures

AITranscriberStoreEvent

AI transcriber event callback.
Warning:
Deprecated: Use transcriberState to observe state changes instead
Enum Value
Description
onReceiveTranscriberMessage
Triggered when a new transcription message is received.
onRealtimeTranscriberStarted
Triggered when the real-time transcription service has been started successfully.
onRealtimeTranscriberStopped
Triggered when the real-time transcription service has been stopped.
onRealtimeTranscriberError
Triggered when an error occurs during real-time transcription.

SourceLanguage

Source language type for transcription recognition.
Enum Value
Value
Description
chineseEnglish
zh_en
Chinese-English mixed recognition.
chinese
zh
Chinese.
english
en
English.
cantonese
zh-yue
Cantonese.
vietnamese
vi
Vietnamese.
japanese
ja
Japanese.
korean
ko
Korean.
indonesian
id
Indonesian.
thai
th
Thai.
portuguese
pt
Portuguese.
turkish
tr
Turkish.
arabic
ar
Arabic.
spanish
es
Spanish.
hindi
hi
Hindi.
french
fr
French.
malay
ms
Malay.
filipino
fil
Filipino.
german
de
German.
italian
it
Italian.
russian
ru
Russian.

TranslationLanguage

Target language type for translation.
Enum Value
Value
Description
chinese
zh
Chinese.
english
en
English.
vietnamese
vi
Vietnamese.
japanese
ja
Japanese.
korean
ko
Korean.
indonesian
id
Indonesian.
thai
th
Thai.
portuguese
pt
Portuguese.
arabic
ar
Arabic.
spanish
es
Spanish.
french
fr
French.
malay
ms
Malay.
german
de
German.
italian
it
Italian.
russian
ru
Russian.

InterpretationVoice

Voice type for interpretation.
Enum Value
Value
Description
female
female
Female voice.
male
male
Male voice.

TranscriberMessage

Transcription message, representing a single segment of transcribed speech.
Property
Type
Description
segmentId
String
Unique identifier of the transcription segment.
speakerUserId
String
User ID of the speaker.
speakerUserName
String
Display name of the speaker.
sourceText
String
Original transcribed text from the speech recognition.
translationTexts
[TranslationLanguage: String]
Translation results, keyed by target language. Contains translation texts for each configured translation language.
timestamp
Int64
Timestamp of the message.
isCompleted
Bool
Whether the segment transcription is completed. When true, the source text and translation texts will no longer be updated.

TranscriptionConfig

Configuration for transcription features.
Property
Type
Description
enableTranslation
Bool
Whether to enable translation during transcription.

InterpretationConfig

Configuration for interpretation features.
Property
Type
Description
voice
Voice type for interpretation output.

TranscriberConfig

Transcription configuration, used to configure source language and translation languages.
Warning:
Deprecated: Use TranscriptionConfig or InterpretationConfig instead
Property
Type
Description
sourceLanguage
Source language for speech recognition, defaults to Chinese-English mixed.
translationLanguages
List of target languages for translation. Multiple languages can be configured simultaneously.

TranscriberState

Transcription state data provided by AITranscriberStore.
Property
Type
Description
selfLanguage
Current user's self language.
realtimeMessageList
List of real-time transcription messages, updated automatically as transcription progresses.
isTranscriptionRunning
Bool
Whether transcription is currently running.
transcriptionConfig
Current transcription configuration.
isInterpretationRunning
Bool
Whether interpretation is currently running.
interpretationConfig
Current interpretation configuration.
interpretationVolume
Int
Current interpretation volume level.

Usage Example

// Create instance for a specific room
let store = AITranscriberStore.create(roomID: "your_room_id")
// Subscribe to state changes
store.state.subscribe { state in
    for message in state.realtimeMessageList {
        print("Speaker: \\(message.speakerUserName)")
        print("Source text: \\(message.sourceText)")
    }
}
// Configure transcription
var config = TranscriberConfig()
config.sourceLanguage = .chineseEnglish
config.translationLanguages = [.english, .japanese]
// Start transcription
store.startRealtimeTranscriber(config: config) { error in
    if let error = error {
        print("Start failed: \\(error)")
    }
}

前の記事へ: Define次の記事へ: AudioEffectStore

ヘルプとサポート

この記事はお役に立ちましたか？

フィードバック