TTS_多國語言語音合成-Sync v3:

TTS Speeches v3

提供語音轉文字，讓您將多國文本藉由 REST API 轉換為合成語音，並且支援自然的的聲調與特定語言，接下來說明如何用 http request 來取得文字轉語音服務。

您可以在這個 Demo 測試網站輸入金鑰進行測試：https://taiwanailabs-yating.github.io/asr-streaming-js-demo/

Host: https://tts.api.yating.tw

送出語音合成需求

在處理完所有輸入後，一般情況下，回應時間低於 120 秒。

Request

URL: /v3/speeches/synchronize
Method: POST

Header

Body

Input設定

Voice設定

Model列表

AudioConfig設定

AudioConfig 中的 Encoding 音訊編碼設定

除了聲音類型，還可以選擇音頻輸出格式。 目前支持 MP3 和 LINEAR16。

Request Body Example

{
    "input": {
        "text": "天下國際週報",
        "type": "text"
    },
    "voice": {
        "model": "female_1",
        "lang": "zh_tw"
    },
    "audioConfig": {
        "encoding": "LINEAR16",
        "maxLength": 600000,
        "uploadFile": false
    }
}

Success Response Body Example("uploadFile": true)

{
    "uid": "484bb379-9c8e-4a37-99dd-4906852edff3",
    "inputText": "天下國際週報",
    "inputType": "text",
    "voiceModel": "female_1",
    "voiceLang": "zh_tw",
    "audioEncoding": "LINEAR16",
    "audioMaxLength": 600000,
    "status": "completed",
    "audioFile": {
        "fileName": "c81405ae-4d93-400a-8093-531cf4913b8f_speed_up.wav",
        "fileHash": "99757fc4f19ae1767b24b254588aa5e79de82b67",
        "url": "https://tts.api.yating.tw/tts-files/tts-public/tts-audio/c81405ae-4d93-400a-8093-531cf4913b8f_speed_up.wav?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=minioadmin%2F20240513%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20240513T095629Z&X-Amz-Expires=86400&X-Amz-SignedHeaders=host&X-Amz-Signature=44b0b9df46dd7077393904afdcbdbb91807805bdecef9295d413c3a6f43fc824",
        "expiredAt": "",
        "errorMessage": ""
    }
}

Success Response Body Example("uploadFile": false)

{
    "uid": "484bb379-9c8e-4a37-99dd-4906852edff3",
    "inputText": "天下國際週報",
    "inputType": "text",
    "voiceModel": "female_1",
    "voiceLang": "zh_tw",
    "audioEncoding": "LINEAR16",
    "audioMaxLength": 600000,
    "status": "completed",
    "audioFile": {
        "fileName": "48bc2337-0b42-461c-8424-802f5c6ca853_speed_up.wav",
        "fileHash": "58a26459a8f3dfe5f8d166b7efa064194e29dae9",
        "url": "",
        "expiredAt": "",
        "errorMessage": "",
        "audioContent": "UklGRjZ5AQBXQVZFZm10IBAAAAABAAEARKwAAIhYAQACABAATElTVBoAAABJTkZPSVNGVA4AAABMYXZmNTguNzYuMTAwAGRhdGHweAEAAQAAAAIABgAIAA0AEgASAAkA/v/x//L//f8HAAgA/...."
    }
}

Failure Response Body Example

{
  statusCode: 400,
  message: string[],
  error: "Bad Request"
}

• [201] created， 可以在audioContent 裡取得 base64 加密後的音檔內容，解密後存入檔案內即可播放。
• [400] invalid request format: request 參數不符合規定，如輸入的聲音編碼不存在，sampleRate不支援
• [401] unauthorized: 密鑰不存在或超出限制。
• [422] pipeline error: 請查看下表（管道錯誤消息）以獲取更多詳細信息。
• [500] internal server error

Pipeline Error message

限制

• 每個key的最大同時需求處理數：3

More voices