We are excited to announce the release of JSVoice 1.0.2. This update brings a massive set of improvements, focusing on flexibility, developer experience, and production readiness.

What is JSVoice?

JSVoice is a production-ready JavaScript library for adding voice commands and speech synthesis to web applications. Built on the Web Speech API, it provides an intuitive interface for creating voice-enabled experiences with zero dependencies.

It is perfect for:

Accessibility: Enabling hands-free navigation.
Voice-First Interfaces: Building conversational UIs.
Productivity Tools: Creating efficiency-focused voice apps.

New in Version 1.0.2

1. Pluggable Speech Engines

This is the biggest architectural change. You are no longer limited to the browser's native Web Speech API (which can be inconsistent across browsers).
JSVoice now supports a BaseSpeechEngine architecture, allowing you to plug in custom engines like OpenAI Whisper, Google Cloud Speech-to-Text, or your own backend solution.

2. Real-Time Audio Visualization

We have added built-in support for visualizing voice input. You can now easily create:

Waveform displays
Frequency bars
Amplitude monitoring

voice.startAmplitude((bars) => {
  // bars is an array of normalized values (0-1)
  updateVisualizer(bars);
}, {
  mode: 'bars',
  barCount: 16
});

3. Advanced Pattern Matching

Commands are now more dynamic than ever. With our new pattern matching engine, you can extract variables directly from speech:

// User says: "set volume to 75"
voice.addPatternCommand('set volume to {level}', (args) => {
  console.log(`Volume set to ${args.level}%`); // { level: "75" }
});

4. Better TypeScript Support

We have completely rewritten our type definitions. JSVoice now provides full type safety for configuration options, events, and engine implementations, giving you a top-tier IDE experience.

Quick Start

Getting started is easier than ever.

npm install jsvoice

import JSVoice from 'jsvoice';

const voice = new JSVoice();

// Add a simple command
voice.addCommand('hello world', () => {
  voice.speak('Hello! How can I help you?');
});

// Start listening
voice.start();

Why JSVoice?

Zero Dependencies: Keeps your bundle size small (32KB minified).
Privacy Focused: Processes data locally where possible (depending on the engine).
Cross-Browser: Normalizes the differences between Chrome, Edge, Safari, and more.

What's Next?

We are working on official adapter packages for popular AI speech services and deeper integration with LLMs for intent understanding.

Check out the GitHub Repository or the full documentation to learn more.