Speech

Intellinode speech synthesis controller enables your application to convert text into spoken words, leveraging advanced text-to-speech (TTS) technologies from leading providers such as Google and OpenAI.

Parameters

To synthesize speech, you'll need to specify:

• provider: The AI service provider ('google' or 'openai').
• apiKey: Your API key for the chosen provider.
• text: The text content you want to convert to speech.
• language (Google only): The language code for the speech output.
• gender (Google only): Selection between FEMALE and MALE.
• model (OpenAI only): Specifies the OpenAI model variant for speech synthesis.
• voice (OpenAI only): The voice model to use for speech output.

For google provider, the language code (language) parameter determines the accent and language of the synthesized speech. Supported languages include English (en-gb or en for British English), Turkish (tr-tr or tr), Mandarin Chinese (cmn-cn or cn), German (de-de or de), and Arabic (ar-xa or ar), among others. The gender (gender) parameter allows for the selection between FEMALE and MALE voices.

Example

const { RemoteSpeechModel, SupportedSpeechModels, Text2SpeechInput } = require('intellinode');

Google Text-to-Speech

const remoteSpeechModel = new RemoteSpeechModel(process.env.GOOGLE_API_KEY, SupportedSpeechModels.GOOGLE);

const input = new Text2SpeechInput({
  text: 'Welcome to IntelliNode',
  language: 'en-gb'
});

// get the audio content
const audioContent = await remoteSpeechModel.generateSpeech(input);

Save the audio:

const AudioHelper = require('intellinode');

const audioHelper = new AudioHelper();
const decodedAudio = audioHelper.decode(audioContent);
const saved = audioHelper.saveAudio(decodedAudio, tempDir, 'temp.mp3');

OpenAI Text-to-Speech

const openAiRemoteSpeechModel = new RemoteSpeechModel(process.env.OPENAI_API_KEY, SupportedSpeechModels.OPENAI);

const input = new Text2SpeechInput({
  model: 'tts-1',
  text: "The quick brown fox jumped over the lazy dog.",
  voice: "alloy",
  stream: true
});

const result = await openAiRemoteSpeechModel.generateSpeech(input);

Save the audio:

const fs = require('fs');

const filePath = './temp/downloaded_audio.mp3';

// create the write stream
const writer = fs.createWriteStream(filePath);
result.pipe(writer);

// handle the completion of writing the file
writer.on('finish', () => {
  const fileExists = fs.existsSync(filePath);
  console.log('Audio file downloaded successfully!');
});