Whisper WebUI

A web-based interface for Whisper Transcribe, providing easy-to-use audio transcription and text enhancement features.

Features

• Transcribe Tab: Convert audio files to text using OpenAI Whisper

  • Audio file upload with preview
  • Start/End time selection with sliders
  • Real-time progress tracking
  • Interactive transcript viewer with audio synchronization
  • Click text segments to jump to audio timestamp
  • Auto-highlight text as audio plays

• Enhance Tab: Improve transcripts using Gemini API

  • Process completed transcriptions
  • Customizable enhancement prompts
  • Queue-based background processing

• Transcript Viewer: Interactive result viewing

  • Timestamp-synced audio player
  • Click any text to jump to that moment in audio
  • Auto-scroll and highlight current segment
  • Download results as Markdown

Tech Stack

• Frontend: Vite + TailwindCSS
• Backend: FastAPI + Python
• AI Models: OpenAI Whisper + Google Gemini
• Infrastructure: Docker Compose

Quick Start

1. Copy the environment template:

   cp .env.example .env

2. Edit .env and add your Gemini API key:

   GEMINI_API_KEY=your-actual-api-key
   

3. Start the services:

   docker-compose up -d

4. Open your browser:

   • Frontend: http://localhost:5173
   • Backend API: http://localhost:8000

Configuration

All configuration is done via the .env file. See .env.example for available options.

Required Configuration

• GEMINI_API_KEY: Your Google Gemini API key (get it from Google AI Studio)

Optional Configuration

AI Models:

• WHISPER_MODEL: Whisper model to use (default: openai/whisper-large-v3-turbo)
• ENABLE_FLASH_ATTENTION: Enable Flash Attention 2 for faster GPU processing (default: false)

Server Settings:

• BACKEND_PORT: Backend server port (default: 8000)
• FRONTEND_PORT: Frontend server port (default: 5173)
• RELOAD_MODE: Enable auto-reload for development (default: true)

Security:

• ALLOWED_ORIGINS: Comma-separated list of allowed CORS origins (default: http://localhost:5173)
  • For development: http://localhost:5173,http://localhost:3000
  • For production: https://yourdomain.com

File Handling:

• MAX_FILE_SIZE: Maximum upload file size (default: 500MB)
• UPLOAD_DIR: Directory for uploaded files (default: /app/uploads)
• OUTPUT_DIR: Directory for output files (default: /app/outputs)

Logging:

• LOG_LEVEL: Logging level - DEBUG, INFO, WARNING, ERROR (default: INFO)
• LOG_FORMAT: Log format - json or text (default: json)

Security Features

This application includes several security enhancements:

• CORS Protection: Configurable allowed origins to prevent unauthorized access
• File Validation:
  • Extension validation for audio files
  • MIME type validation using magic bytes
  • File size limits
• Path Traversal Prevention: Filename sanitization to prevent directory traversal attacks
• Unique File Naming: UUID-based filenames to prevent collisions and overwrites
• Thread-Safe Operations: Singleton services with proper locking mechanisms
• Structured Logging: Comprehensive logging for monitoring and debugging

Development

Backend Development

cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
uvicorn main:app --reload

Frontend Development

cd frontend
npm install
npm run dev

System Requirements

• Docker & Docker Compose
• 4GB+ RAM
• 10GB+ disk space (for models and data)
• Optional: CUDA-compatible GPU for faster processing

License

MIT