deepcast

deepcast is an AI-powered CLI tool that generates engaging podcast-style conversations with realistic text-to-speech capabilities. Perfect for creating educational content, practice conversations, or exploring topics in a dialogue format.

✨ Features

🤖 AI-Powered Conversations: Uses Deepseek-V3 model for generating natural, educational dialogues
🎧 Interactive Format: Generates engaging podcast-style conversations between two speakers
📚 Educational Content: Creates deep, insightful discussions on any given topic
🗣️ Text-to-Speech: Integrates PlayHT for converting conversations into realistic audio
🚀 Background Music: Add ambient music with adjustable volume
😊 Voice Emotions: Control speaker emotions (happy, serious, excited, etc.)
📄 Rich File Support: Generate from TXT, PDF, DOCX, EPUB, Markdown, HTML files
🌐 Web Content: Generate from web articles, YouTube transcripts, and URLs
🔄 Content Combination: Combine multiple sources into one podcast
🌍 Multiple Languages: Support for English, Spanish, French, German, Italian, and Portuguese
🎭 Podcast Styles: Different conversation styles (interview, debate, storytelling, etc.)
📊 Complexity Levels: Adjust content for beginner, intermediate, or expert audiences
🚀 Easy to Use: Simple CLI interface with rich terminal output

🛠️ Installation

Clone the repository:

git clone https://github.com/byigitt/deepcast.git
cd deepcast

Install dependencies using uv:

uv venv
uv pip install -e .

Create a .env file from the example:

cp .env.example .env

Add your API keys to the .env file:

Get an OpenRouter API key from OpenRouter
Get a FAL API key from FAL.ai

🚀 Usage

View Available Options

List available podcast styles:

deepcast styles

List available background music:

deepcast music

List available voice emotions:

deepcast emotions

Generate from a Topic

Create a podcast about any topic with custom settings:

# Basic usage
deepcast generate "Quantum Computing"

# With custom style
deepcast generate "Quantum Computing" --style debate

# With background music
deepcast generate "Quantum Computing" --music ambient --volume 0.2

# With voice emotions
deepcast generate "Quantum Computing" \
    --speaker1-emotion professional \
    --speaker2-emotion friendly

# Full customization
deepcast generate "Quantum Computing" \
    --style educational \
    --complexity expert \
    --language french \
    --exchanges 7 \
    --music soft_piano \
    --volume 0.15 \
    --speaker1-emotion serious \
    --speaker2-emotion excited \
    --save-audio \
    --format mp3

Generate from Files

Create a podcast from various file types:

# From a single file with music
deepcast generate "Research Paper" \
    --file paper.pdf \
    --music ambient

# From multiple files with emotions
deepcast generate "Research Summary" \
    --file paper1.pdf \
    --file paper2.pdf \
    --speaker1-emotion professional \
    --speaker2-emotion friendly

# From different file types with full audio
deepcast generate "Documentation" \
    --file intro.md \
    --file chapter1.docx \
    --file appendix.pdf \
    --music soft_piano \
    --volume 0.2 \
    --save-audio

Generate from Web Content

Create a podcast from web content:

# From a web article with music
deepcast generate "News Article" \
    --url "https://example.com/article" \
    --music cinematic

# From a YouTube video with emotions
deepcast generate "Video Summary" \
    --youtube "https://youtube.com/watch?v=..." \
    --speaker1-emotion excited \
    --speaker2-emotion professional

# Combine web and file content with full audio
deepcast generate "Research Review" \
    --file paper.pdf \
    --url "https://example.com/article" \
    --youtube "https://youtube.com/watch?v=..." \
    --music jazz \
    --volume 0.1 \
    --save-audio

Output Options

Save the transcript to a file:

deepcast generate "Artificial Intelligence" --output transcript.txt

Only get the audio URL:

deepcast generate "Space Exploration" --audio-only

Save audio locally:

deepcast generate "Nature Documentary" \
    --music nature \
    --save-audio \
    --format mp3

Full Example

Combine all features:

deepcast generate "Advanced Physics" \
    --file research.pdf \
    --file notes.md \
    --url "https://example.com/article" \
    --youtube "https://youtube.com/watch?v=..." \
    --style educational \
    --complexity expert \
    --language french \
    --exchanges 7 \
    --music cinematic \
    --volume 0.15 \
    --speaker1-emotion professional \
    --speaker2-emotion excited \
    --save-audio \
    --format mp3 \
    --output transcript.txt

🏗️ Project Structure

src/
├── models/         # Data models (Podcast, Config, Audio)
├── services/       # Core services (LLM, Audio, File, Content)
├── utils/          # Utility functions (Config)
└── cli.py          # CLI interface

🔧 Configuration

The following environment variables can be configured in .env:

OPENROUTER_API_KEY: Your OpenRouter API key for accessing the Deepseek model
FAL_KEY: Your FAL.ai API key for text-to-speech conversion
LOG_LEVEL: Optional logging level (default: INFO)

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

Fork the repository
Create your feature branch (git checkout -b feature/AmazingFeature)
Commit your changes (git commit -m 'Add some AmazingFeature')
Push to the branch (git push origin feature/AmazingFeature)
Open a Pull Request

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

OpenRouter for providing access to the Deepseek model
FAL.ai for the text-to-speech capabilities
PlayHT for voice synthesis
Pixabay for background music
All our contributors and users

License

EudaLabs/deepcast

ErrorLooks like something went wrong!

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Languages

Error
Looks like something went wrong!