whisper-nodejs

whisper-nodejs is an npm package for using OpenAI's Whisper API to transcribe and translate audio. With whisper-nodejs, you can easily convert audio files into text and translate them into English or other supported languages.

Installation

Install with npm:

npm install whisper-nodejs

Usage

const Whisper = require('whisper-nodejs');
const whisper = new Whisper('YOUR_API_KEY');

// Transcribe audio
whisper.transcribe('/path/to/file/openai.mp3', 'whisper-1')
  .then(text => {
    console.log(text);
  })
  .catch(error => {
    console.error(error);
  });

// Translate audio
whisper.translate('/path/to/file/german.mp3', 'whisper-1', 'en')
  .then(text => {
    console.log(text);
  })
  .catch(error => {
    console.error(error);
  });

API

new Whisper(apiKey)

Creates a new Whisper instance for calling the transcribe and translate methods.

const whisper = new Whisper('YOUR_API_KEY');

apiKey: Your OpenAI API key.

whisper.transcribe(filePath, modelName)

Transcribes an audio file to text.

whisper.transcribe('/path/to/file/openai.mp3', 'whisper-1')
  .then(text => {
    console.log(text);
  })
  .catch(error => {
    console.error(error);
  });

filePath: The path to the audio file to transcribe. modelName: The name of the Whisper model.

whisper.translate(filePath, modelName, targetLanguage)

Translates an audio file into English or another supported language.

whisper.translate('/path/to/file/german.mp3', 'whisper-1', 'en')
  .then(text => {
    console.log(text);
  })
  .catch(error => {
    console.error(error);
  });

filePath: The path to the audio file to translate. modelName: The name of the Whisper model. targetLanguage: The target language code, such as en for English, zh for Chinese, and so on.

References

https://platform.openai.com/docs/api-reference/introduction