Speak Easy TTS 🎤✨

Welcome to Speak Easy TTS – your playful, powerful, and easy-to-use text-to-speech engine powered by Gemini AI! This package is designed for a top-tier developer experience with a modern, fully-typed API, robust error handling, and fun, expressive features.

Features

• 🚀 Modern TypeScript API: Fully-typed and designed for an ergonomic developer experience.
• 🎭 Expressive Voices: A whole cast of fun voices to bring your text to life.
• 🌍 Multilingual Support: Synthesize speech in numerous languages.
• 🧑‍💻 Developer-Friendly Docs: Comprehensive JSDoc and clear examples.
• 🎉 Playful & Powerful: Fun error messages, helpful logs, and robust features like caching and a CLI.
• 🔥 Everything Exported: All the types and helpers you need are ready to import.

Table of Contents

• Installation
• Quick Start
• CLI Usage
• Library Usage
  • Basic Synthesis
  • Using a Random Voice
  • Synthesizing to a Buffer or Stream
  • Using the Cache
  • Multimodal Synthesis
  • Advanced Configuration
• API Reference
• Contributing
• License

Installation

For the CLI:

npm install -g @base83/speak-easy

For the library:

npm install @base83/speak-easy

Quick Start

# Ensure your GEMINI_API_KEY is set in your environment
export GEMINI_API_KEY="YOUR_API_KEY"

# Run from the CLI
speak-easy "Hello from Speak Easy!"

CLI Usage

From the Command Line (CLI)

# Basic synthesis
speak-easy "Hello from the CLI!"

# With options
speak-easy "Bonjour le monde!" --voice Puck --language fr-FR --output bonjour.mp3

# Using cache for repeated requests
speak-easy "This will be cached" --use-cache

As a Library

import { TtsEngine } from '@base83/speak-easy';
const engine = new TtsEngine({ apiKey: process.env.GEMINI_API_KEY, debug: true });
await engine.synthesizeToFile({ text: 'Hello world!' });

Library Usage

Basic Synthesis

import { TtsEngine } from '@base83/speak-easy';
const engine = new TtsEngine({ apiKey: process.env.GEMINI_API_KEY });
const { filePath } = await engine.synthesizeToFile({ text: 'Hello from the library!' });
console.log(`Audio saved to: ${filePath}`);

Using a Random Voice

import { TtsEngine, getRandomVoice } from '@base83/speak-easy';
const engine = new TtsEngine({ apiKey: process.env.GEMINI_API_KEY });
const voice = getRandomVoice();
await engine.synthesizeToFile({ text: 'A random voice is fun!', voice });

Synthesizing to a Buffer or Stream

// To Buffer
const { buffer } = await engine.synthesizeToBuffer({ text: 'This is a buffer.' });

// To Stream
const { stream } = await engine.synthesizeToStream({ text: 'This is a stream.' });
stream.pipe(process.stdout);

Using the Cache

Enable caching to save time and API calls on repeated requests.

const engine = new TtsEngine({ apiKey: process.env.GEMINI_API_KEY, useCache: true });
await engine.synthesizeToFile({ text: 'This will be cached.' });
await engine.synthesizeToFile({ text: 'This will be cached.' }); // This one will be much faster!

Multimodal Synthesis

Use files as prompts for synthesis.

const fileMeta = await engine.uploadFile({ file: 'image.png', mimeType: 'image/png' });
await engine.synthesizeToFile({ file: fileMeta, prompt: 'Describe this image for me.' });

Advanced Configuration

Customize the engine with hooks, a default voice, and your own logger.

const engine = new TtsEngine({
  apiKey: process.env.GEMINI_API_KEY,
  defaultVoice: 'Puck',
  useCache: true,
  hooks: {
    beforeSynthesize: (opts) => console.log('Starting job...'),
    afterSynthesize: (res) => console.log(`Finished with ${res.voice}!`),
  }
});

API Reference

• TtsEngine: Main class for all TTS magic
• getRandomVoice(): Get a random voice for fun variety
• GEMINI_TTS_VOICES: List of all available voices
• GEMINI_TTS_LANGUAGES: List of supported languages
• isValidVoice() / isValidLanguage(): Validate your choices

Error Handling

All errors are playful and descriptive! For example:

🛑 Oops! No API key found for your TTS adventure.
Set GEMINI_API_KEY in your environment or pass it to the TtsEngine constructor. The magic won't work without it! 🪄

Pro Tips

• Use the debug: true option for extra fun logs.
• The synthesizeToFile method will auto-generate a filename if you don't provide one.
• Check out the full JSDoc in your editor for detailed info on every function and type.

Contributing

PRs and ideas welcome! Let's make TTS more fun for everyone.

Go forth and synthesize with style! 🚀