Skip to main content

Python

This sample script demonstrates how to interact with the Whisperly TTS API to convert text into speech. The script sends a POST request to the /generate-voice endpoint with the required form data, including an optional speaker sample file, and prints the JSON response.

API Endpoints​

POST /transcription

  • URL: https://tts.recordly.ai/generate-voice
  • Form Data Parameters:
    • text (required): The text that you want to convert into speech.
    • speaker_wav: Used to upload a speaker sample WAV file. This file can guide the voice synthesis to mimic a particular speaker’s tone or style.
    • language: Specifies the language of the generated voice (default is "tr").
    • file_path: The name for the output audio file (default is "output.wav").
    • speed: Controls the speed of the voice output (default is "2.0").
    • split_sentences: Determines whether the text should be split into sentences before processing (default is "false").
  • Making the Request:
    • The script uses the requests library to send a multipart/form-data POST request. If the request is successful, the JSON response containing details about the generated audio file is printed to the console.

Usage​

The example below shows how to start a text-to-speech request and then retrieve the result.

import requests
import json

BASE_URL = "https://tts.recordly.ai"
TOKEN = "<TOKEN>"

def generate_voice(text: str, speaker_wav_path: str, language: str = "en",
                   file_path: str = "output.wav", speed: str = "2.0", split_sentences: str = "false"):
    """
    Generates voice from text using the Whisperly TTS service.

    Parameters:
    -----------
    text : str
        The text to be converted into speech.
    speaker_wav_path : str
        The file path to the speaker sample WAV file.
    language : str, optional
        The language code for the voice output. Default is "tr".
    file_path : str, optional
        The desired output filename for the generated audio. Default is "output.wav".
    speed : str, optional
        The speed factor for the voice output. Default is "2.0".
    split_sentences : str, optional
        Whether to split the text into sentences before processing. Default is "false".

    Returns:
    --------
    audio file
        The generated audio file.
    """
    endpoint = f"{BASE_URL}/generate-voice"
    headers = {
        "Authorization": f"Bearer {TOKEN}"
    }
    data = {
        "text": text,
        "language": language,
        "file_path": file_path,
        "speed": speed,
        "split_sentences": split_sentences
    }
    files = {
        "speaker_wav": open(speaker_wav_path, "rb")
    }
    
    print("Sending voice generation request with the following payload:")
    print(json.dumps(data, indent=2))
    
    response = requests.post(endpoint, headers=headers, data=data, files=files)
    response.raise_for_status()
    return response.json()

if __name__ == "__main__":
    # The text to be converted into speech
    text_to_convert = "Lorem ipsum dolor sit amet, consectetur adipiscing elit."
    
    # Path to the speaker sample WAV file (adjust the path as needed)
    speaker_wav_path = "/path/to/file/speaker.wav"
    
    print("Starting voice generation...")
    result = generate_voice(text=text_to_convert, speaker_wav_path=speaker_wav_path)
    
    print(f"Voice generation complete. Audio saved to 'output.wav'")