ai-text-to-speech

ai-text-to-speech

A powerful and straightforward Node.js module for generating speech audio from text using the OpenAI API (support for other TTS providers in the works). ai-text-to-speech offers a simple and robust interface to convert text into high-quality speech audio files in various formats and voices.

Developed by Jerry Kapron for everyone to use as they please.
☕️ Buy me a coffee

Table of Contents

• ai-text-to-speech - ☕️ Buy me a coffee
  • Table of Contents
  • Features
  • Installation
  • Usage
    • Basic Example
    • Advanced Usage
  • Options
    • input (string, required)
    • dest_dir (string, optional)
    • file_name (string, optional)
    • voice (string, optional)
    • model (string, optional)
    • response_format (string, optional)
    • suffix_type (string, optional)
    • api_key (string, optional)
  • Allowed Values
    • Voices
    • Models
    • Response Formats
    • Suffix Types
  • Edge Cases and Warnings
  • License
  • Acknowledgements
    • Audio Format Descriptions:

Features

• Easy Integration: Seamlessly integrate text-to-speech functionality into your Node.js applications.
• Multiple Voices: Choose from a variety of high-quality voices to suit your application's needs.
• Flexible Output Formats: Supports various audio formats like MP3, WAV, FLAC, and more.
• Customizable File Naming: Control the output file naming with suffix options to prevent overwrites.
• Robust Error Handling: Comprehensive validation and descriptive error messages for easy debugging.

Installation

Install ai-text-to-speech via NPM:

npm install ai-text-to-speech

Usage

How to load the module in your code

// Use this import statement if your project supports ES Modules
import aiSpeech from 'ai-text-to-speech';

OR

// Use this require statement if your project uses CommonJS modules
const aiSpeech = require('ai-text-to-speech');

Basic Example with async/await

(async () => { // or nested inside another async function

  try {
    const audioFilePath = await aiSpeech({
      input: 'Buy me a coffee if it works for you.'
      // If the OPENAI_API_KEY environment variable is already set,
      // you don't have to specify the api_key option
    });
    console.log(`Audio file saved at: ${audioFilePath}`);
  } catch (error) {
    console.error('Error generating speech audio:', error.message);
  }

})();

Basic Example with Promise .then()/.catch()

// This approach can be useful if you prefer working with promises directly
// or if you're in an environment where async/await is not supported.

aiSpeech({
  input: 'Buy me a coffee if it works for you.',
  // You can explicitly provide your OpenAI API key here
  api_key: 'YOUR_OPENAI_API_KEY', // if process.env.OPENAI_API_KEY is not set
})
.then((audioFilePath) => {
    console.log(`Audio file saved at: ${audioFilePath}`);
})
.catch((error) => {
    console.error('Error generating speech audio:', error.message);
});

Advanced Usage with async/await

(async () => { // or nested inside another async function

  try {
    const audioFilePath = await aiSpeech({
      input: 'Buy me a coffee if it works for you.',
      dest_dir: './audio',
      file_name: 'welcome-message',
      voice: 'echo',
      model: 'tts-1-hd',
      response_format: 'wav',
      suffix_type: 'nano',
      api_key: 'YOUR_OPENAI_API_KEY', // if process.env.OPENAI_API_KEY is not set
    });
    console.log(`Audio file saved at: ${audioFilePath}`);
  } catch (error) {
    console.error('Error generating speech audio:', error.message);
  }

})();

Advanced Usage with Promise .then()/.catch()

// This approach can be useful if you prefer working with promises directly
// or if you're in an environment where async/await is not supported.

aiSpeech({
    input: 'Buy me a coffee if it works for you.',
    dest_dir: './audio',
    file_name: 'welcome-message',
    voice: 'echo',
    model: 'tts-1-hd',
    response_format: 'wav',
    suffix_type: 'nano',
    api_key: 'YOUR_OPENAI_API_KEY', // if process.env.OPENAI_API_KEY is not set
})
.then((audioFilePath) => {
    console.log(`Audio file saved at: ${audioFilePath}`);
})
.catch((error) => {
    console.error('Error generating speech audio:', error.message);
});

Options

input (string, required)

The text to generate audio for. Maximum length is 4096 characters.

dest_dir (string, optional)

The destination directory to save the audio file. Default: './' (current directory).

file_name (string, optional)

The base name of the output file. Default: 'speech-audio'.

voice (string, optional)

The voice to use for speech synthesis. Default: 'nova'.

model (string, optional)

The TTS model to use. Default: 'tts-1'.

response_format (string, optional)

The audio format for the output file. Default: 'mp3'.

suffix_type (string, optional)

The type of unique suffix used in the file name. Default: 'uuid'.

api_key (string, optional)

Your OpenAI API key. Default: The value of the OPENAI_API_KEY environment variable.

Allowed Values

Voices

• alloy
• echo
• fable
• onyx
• nova (default)
• shimmer

Models

• tts-1 (default)
• tts-1-hd

Response Formats

• mp3 (default)
• opus
• aac
• flac
• wav
• pcm

Suffix Types

• uuid (default): A unique UUID string.
• milli: Timestamp in milliseconds.
• micro: Timestamp in microseconds.
• nano: Timestamp in nanoseconds.
• none: No suffix. Warning: May overwrite existing files if filenames collide.

Edge Cases and Warnings

• Input Length: The input text must not exceed 4096 characters. Exceeding this limit will result in an error.
• File Overwrite Risk: Using suffix_type: 'none' without specifying a unique file_name may lead to overwriting existing files.
• Directory Permissions: Ensure the dest_dir exists and the application has write permissions. The module will throw an error if it cannot write to the directory.
• API Key Requirement: An OpenAI API key is required. Set it via the api_key option or the OPENAI_API_KEY environment variable.
• Network Errors: Network issues or incorrect API endpoints will result in errors. Ensure you have a stable internet connection.
• Unsupported Values: Providing unsupported values for voice, model, response_format, or suffix_type will result in an error.

License

This project is licensed under the MIT License.

Acknowledgements

• OpenAI for providing the API used in this module.
• Node.js community for their continuous support.

---

For more information on voice options, see the OpenAI Text-to-Speech Voice Options.

Audio Format Descriptions:

• Opus: Ideal for internet streaming and communication due to low latency.
• AAC: Preferred for digital audio compression; widely used on YouTube, Android, and iOS.
• FLAC: Suitable for lossless audio compression; favored by audio enthusiasts for archiving.
• WAV: Uncompressed audio, suitable for applications requiring minimal decoding overhead.
• PCM: Raw audio samples in 24kHz (16-bit signed, little-endian), without headers.

---

Note: Ensure compliance with OpenAI's usage policies when integrating this module into your applications.

---

☕️ Buy me a coffee

---

Back to top