PPTX Creator

A command-line tool to convert Markdown content into professional PowerPoint presentations using AI-powered layout selection.

Features

• Convert Markdown to PowerPoint presentations.
• AI-powered content analysis and layout selection (OpenRouter LLM).
• Project-based workflow: each presentation is a project directory.
• Custom PowerPoint template support.
• Dual image support:
  • Manual: Include your own images via Markdown syntax
  • AI-Generated: Automatic image generation via Stable Diffusion Forge
• Extensible and testable Python codebase.

Requirements

• Python 3.10+
• pip
• An OpenRouter API key (set as OPENROUTER_API_KEY in a .env file)
• Git (for cloning)

Installation

1. Clone the repository:

   git clone https://github.com/your-username/pptx-creator.git
   cd pptx-creator

2. Create and activate a virtual environment:

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt
   # For development:
   pip install -r requirements-dev.txt

4. Set up your API key:

   • Create a file named .env in the project root:

     OPENROUTER_API_KEY=your-openrouter-api-key-here
     

Usage

The tool uses a project-based approach. Each presentation lives in its own directory under projects/.

Image Handling

The tool supports two ways to include images in your presentations:

1. Manual Image Inclusion

   • Place images in your project's images/ directory
   • Reference them in your Markdown using standard syntax:

     ![Alt text](images/your_image.png)

2. AI Image Generation (Optional)

   • The system can automatically generate relevant images using Stable Diffusion Forge
   • During content processing, the AI analyzes your content and may generate supporting images
   • Generated images are saved to your project's images/ directory
   • Requires Stable Diffusion Forge setup (see Configuration below)

1. Prepare Your PowerPoint Template

• Place your .pptx template in the templates/ directory (e.g., templates/my_template.pptx).
• Ensure slide layouts have clearly named placeholders for best results.

2. Create a Project and Add Content

• Create a new directory under projects/ (e.g., projects/robotics/).
• Add your Markdown content as content.md in the project directory.
• Place any images in projects/<project_name>/images/ and reference them in your Markdown:
  ![Alt text](images/my_image.png)

3. Analyze the Template

python -m pptx_generator.main analyze <project_name> <template_path>

• Example:

  python -m pptx_generator.main analyze robotics templates/my_template.pptx

This generates projects/<project_name>/output/layouts.json describing available layouts and placeholders.

4. Process the Markdown Content

python -m pptx_generator.main process <project_name>

• Example:

  python -m pptx_generator.main process robotics

This generates:

• projects/<project_name>/output/presentation.json (slide plan)
• projects/<project_name>/<project_name>.pptx (final PowerPoint)

Configuration

Create a file named .env in the project root with:

# Required for content processing
OPENROUTER_API_KEY=your-openrouter-api-key-here
OPENROUTER_MODEL_NAME=deepseek/deepseek-chat-v3-0324:free  # Optional, this is the default

# Optional: for AI image generation
SD_FORGE_SERVER_URL=http://your-sd-forge-server:7860
FLUX_MODEL_NAME=black-forest-labs/FLUX.1-schnell  # Optional, this is the default

File Structure

pptx-creator/
├── pptx_generator/
│   ├── __main__.py
│   ├── analyzer.py
│   ├── generator.py
│   ├── main.py
│   ├── models.py
│   ├── processor.py
│   └── utils.py
├── projects/
│   └── <project_name>/
│       ├── content.md
│       ├── images/
│       └── output/
├── templates/
│   └── <your_template>.pptx
├── scripts/
├── tests/
├── requirements.txt
├── requirements-dev.txt
├── .env
└── README.md

Testing

Run all tests with:

pytest

Contributing

1. Fork the repository.
2. Create a feature branch (git checkout -b feature/YourFeature).
3. Commit your changes.
4. Push to your branch.
5. Open a Pull Request.

License

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