Claude Yap transforms PDFs and research topics into engaging, multi-voice podcasts using Claude AI and advanced text-to-speech technology. The app creates natural-sounding conversations in various formats, with custom voices for each speaker and automatically-generated cover art.

• Intelligent PDF Analysis: Upload PDFs and Claude will analyze, summarize, and transform them into conversational podcasts
• Multiple Podcast Formats:
  • Podcast Style: Traditional host/guest conversational format
  • Debate Style: Two perspectives debating topics from the source material
  • Duck Mode: Educational teacher/student dialogue format
• Multi-Speaker Natural Voices: Uses Cartesia TTS API to generate distinctly different voices for each speaker
• Automated Cover Art: Creates custom podcast covers using Claude's artifact generation or Stable Diffusion
• Advanced Audio Player: Professional-grade player with keyboard shortcuts, playback speed control, and progress tracking
• Video Conference Integration: Discuss podcasts via video call through Tavus API integration
• Real-Time Progress Tracking: Monitor podcast generation progress with detailed status updates

• Node.js (v16+)
• Python 3.9+
• Claude API key (Anthropic)
• Cartesia API key (for text-to-speech)
• Hugging Face API token (for cover art generation)
• Tavus API credentials (for video conference)

1. Clone the repository

git clone https://github.com/yourusername/claude-yap.git
cd claude-yap

2. Set up the backend

cd backend
pip install -r requirements.txt
# Create .env file using example template
cp .env.example .env
# Fill in your API keys in the .env file

3. Set up the frontend

cd ../frontend
npm install

1. Start the backend server

cd backend
python app.py

2. Start the frontend development server

cd ../frontend
npm run dev

3. Open your browser and navigate to http://localhost:5173

• React with React Router
• Framer Motion for animations
• Advanced HTML5 audio player

• Python with Flask
• Claude 3.7 Sonnet and Claude 3 Opus for content generation
• Cartesia API for high-quality multi-speaker text-to-speech
• Claude artifacts tool and Hugging Face Stable Diffusion for cover art
• Tavus API for video conference integration
• FFmpeg for audio processing

1. Upload Content: Submit PDFs for analysis or provide a research topic
2. Select Format: Choose between podcast, debate, or educational "duck" mode
3. Script Generation: Claude analyzes content and creates a conversational script
4. Voice Assignment: The system automatically assigns distinct voices to each speaker
5. Audio Generation: Cartesia TTS converts the script into natural-sounding multi-voice audio
6. Cover Art Creation: Automatically generates custom cover art for each podcast
7. Playback & Sharing: Listen to your podcast with advanced playback controls
8. AI Video Discussion: Have 1-on-1 video calls with AI personalities to discuss your podcasts

• Keyboard shortcuts for playback control (space, arrows, number keys)
• Variable playback speed (0.5x to 2x)
• Jump forward/backward functionality

Discuss podcasts with AI personalities through Tavus API video conference integration

• Track listened/unlistened podcasts
• Rename and organize your podcast collection

• Frontend: React SPA with responsive UI and advanced audio player
• Backend: Flask API with Claude integration, audio processing, and data management

• Claude Integration: Uses both Claude 3.7 Sonnet and Claude 3 Opus for different tasks
• Script Processing: Intelligently parses scripts to identify speakers and dialogue
• TTS Engine: Uses Cartesia's advanced TTS API for natural multi-speaker audio
• Cover Art Generation: Uses either Claude artifacts (SVG) or Stable Diffusion (PNG)