ai.z.openapi.api.audio

Interface AudioApi

public interface AudioApi

Audio API for advanced speech processing capabilities Powered by GLM-4-Voice for end-to-end speech understanding and generation Provides text-to-speech, speech-to-text, voice customization, and real-time conversation Supports emotion control, tone adjustment, speed variation, and dialect generation Features real-time audio processing, voice cloning, and multilingual ASR capabilities

Method Summary

Modifier and Type	Method and Description
io.reactivex.rxjava3.core.Single<okhttp3.ResponseBody>	audioCustomization(Map<String,okhttp3.RequestBody> request, okhttp3.MultipartBody.Part voiceData) Voice cloning and customization using advanced neural models Creates custom voice models from provided audio samples with high fidelity Enables personalized speech synthesis preserving unique voice characteristics and speaking style Supports fine-tuning of voice parameters including pitch, timbre, and speaking patterns
io.reactivex.rxjava3.core.Single<okhttp3.ResponseBody>	audioSpeech(AudioSpeechRequest request) Text-to-Speech (TTS) conversion using GLM-4-Voice Converts text input into natural-sounding speech audio with emotion and tone control Supports multiple voices, languages, speed adjustment, and various audio formats Features advanced voice synthesis with customizable emotional expressions and dialects
retrofit2.Call<okhttp3.ResponseBody>	audioSpeechStreaming(AudioSpeechRequest request) Text-to-Speech (TTS) conversion using GLM-4-Voice Converts text input into natural-sounding speech audio with emotion and tone control Supports multiple voices, languages, speed adjustment, and various audio formats Features advanced voice synthesis with customizable emotional expressions and dialects
io.reactivex.rxjava3.core.Single<AudioTranscriptionResult>	audioTranscription(Map<String,okhttp3.RequestBody> request, okhttp3.MultipartBody.Part file) Speech-to-text transcription using GLM ASR models Converts audio files to text with high accuracy and multilingual support Features advanced noise reduction, speaker recognition, and punctuation restoration Supports multiple languages with automatic language detection capabilities
retrofit2.Call<okhttp3.ResponseBody>	audioTranscriptionStream(Map<String,okhttp3.RequestBody> request, okhttp3.MultipartBody.Part file) Streaming speech-to-text transcription using GLM ASR Converts audio files to text with real-time streaming results and low latency Returns transcription results incrementally as they become available for immediate processing Optimized for real-time applications requiring immediate text feedback

Method Detail

audioSpeech

@POST(value="audio/speech")
io.reactivex.rxjava3.core.Single<okhttp3.ResponseBody> audioSpeech(@Body
                                                                                               AudioSpeechRequest request)

Text-to-Speech (TTS) conversion using GLM-4-Voice Converts text input into natural-sounding speech audio with emotion and tone control Supports multiple voices, languages, speed adjustment, and various audio formats Features advanced voice synthesis with customizable emotional expressions and dialects

Parameters:

request - TTS parameters including text, voice selection, emotion, speed, tone, and output format

Returns:

Generated high-quality audio content in specified format with natural prosody

audioSpeechStreaming

@Streaming
 @POST(value="audio/speech")
retrofit2.Call<okhttp3.ResponseBody> audioSpeechStreaming(@Body
                                                                                                  AudioSpeechRequest request)

Text-to-Speech (TTS) conversion using GLM-4-Voice Converts text input into natural-sounding speech audio with emotion and tone control Supports multiple voices, languages, speed adjustment, and various audio formats Features advanced voice synthesis with customizable emotional expressions and dialects

Parameters:

request - TTS parameters including text, voice selection, emotion, speed, tone, and output format

Returns:

Generated high-quality audio streaming in specified format with natural prosody

audioCustomization

@Multipart
 @POST(value="audio/customization")
io.reactivex.rxjava3.core.Single<okhttp3.ResponseBody> audioCustomization(@PartMap
                                                                                                                         Map<String,okhttp3.RequestBody> request,
                                                                                                                         @Part
                                                                                                                         okhttp3.MultipartBody.Part voiceData)

Voice cloning and customization using advanced neural models Creates custom voice models from provided audio samples with high fidelity Enables personalized speech synthesis preserving unique voice characteristics and speaking style Supports fine-tuning of voice parameters including pitch, timbre, and speaking patterns

Parameters:

request - Voice customization parameters including model settings and training options

voiceData - High-quality audio file containing voice samples for cloning (recommended: clear, diverse samples)

Returns:

Voice model creation result with customization status and model performance metrics

audioTranscriptionStream

@Streaming
 @POST(value="audio/transcriptions")
 @Multipart
retrofit2.Call<okhttp3.ResponseBody> audioTranscriptionStream(@PartMap
                                                                                                                          Map<String,okhttp3.RequestBody> request,
                                                                                                                          @Part
                                                                                                                          okhttp3.MultipartBody.Part file)

Streaming speech-to-text transcription using GLM ASR Converts audio files to text with real-time streaming results and low latency Returns transcription results incrementally as they become available for immediate processing Optimized for real-time applications requiring immediate text feedback

Parameters:

request - Transcription parameters including language detection, model selection, and streaming settings

file - Audio file to be transcribed (supports various formats: wav, mp3, m4a, etc.)

Returns:

Streaming transcription results with timestamps and confidence scores

audioTranscription

@POST(value="audio/transcriptions")
 @Multipart
io.reactivex.rxjava3.core.Single<AudioTranscriptionResult> audioTranscription(@PartMap
                                                                                                                              Map<String,okhttp3.RequestBody> request,
                                                                                                                              @Part
                                                                                                                              okhttp3.MultipartBody.Part file)

Speech-to-text transcription using GLM ASR models Converts audio files to text with high accuracy and multilingual support Features advanced noise reduction, speaker recognition, and punctuation restoration Supports multiple languages with automatic language detection capabilities

Parameters:

request - Transcription parameters including language preference, model selection, and output format options

file - Audio file to be transcribed (supports wav, mp3, m4a, flac, and other common formats)

Returns:

Complete transcription result with text, timestamps, confidence scores, and speaker information