Vorlesen

Text-to-speech tool for authors to proofread their work by listening back. Powered by Google Cloud Text-to-Speech.

Features

• 🎯 Paste unlimited text (automatically chunked into 4000 char segments)
• 🎙️ 3 high-quality voices (WaveNet & Neural2)
• 🎵 Adjustable playback speed (0.5x - 2x)
• 💾 Browser localStorage persistence
• ⚡ Progress tracking with retry logic
• 📥 Download as MP3

Setup

1. Install Dependencies

npm install

2. Configure Google Cloud TTS

You need Google Cloud Text-to-Speech API credentials. Choose ONE option:

Option A: API Key (Simplest)

1. Go to Google Cloud Console
2. Enable Text-to-Speech API
3. Go to APIs & Services → Credentials
4. Click Create Credentials → API Key
5. Restrict to Text-to-Speech API only
6. Create .env.local:

   GOOGLE_CLOUD_TTS_API_KEY=your_api_key_here
   

Option B: Service Account (Local Development)

1. Download service account JSON from Google Cloud Console
2. Create .env.local:

   GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json
   

Option C: Service Account (Vercel Deployment)

1. Copy entire contents of service-account.json
2. In Vercel: Project Settings → Environment Variables
3. Add variable:

   Name: GOOGLE_CLOUD_SERVICE_ACCOUNT
   Value: {"type":"service_account","project_id":"...",...}
   

3. Run Development Server

npm run dev

Open http://localhost:3000

Deploy to Vercel

1. Click "Deploy" button above
2. Add environment variable: GOOGLE_CLOUD_SERVICE_ACCOUNT with your service account JSON
3. Deploy!

Tech Stack

• Framework: Next.js 15 (App Router)
• UI: ShadCN UI + Tailwind CSS
• TTS: Google Cloud Text-to-Speech API
• Audio: Web Audio API + lamejs for MP3 encoding
• Storage: Browser localStorage (5MB with LRU eviction)

Architecture

• Text Chunking: Splits text at sentence boundaries (4000 char max)
• Parallel Processing: Converts chunks concurrently with progress tracking
• Audio Stitching: Decodes MP3 → concatenates PCM → re-encodes MP3
• Error Handling: Exponential backoff for rate limits, retry failed chunks
• Cost: $0.004 per 1K characters (75x cheaper than ElevenLabs)

Development

# Run tests
npm run test

# Run E2E tests
npm run test:e2e

# Build for production
npm run build

License

MIT