PowerPoint to Video Converter (Legacy)

A full-stack web application that converts PowerPoint presentations into narrated videos using AI-powered script generation and text-to-speech synthesis.

Overview

This application accepts a .pptx file, uses Google Gemini to generate a natural presenter script for each slide, synthesizes speech with Coqui TTS, and assembles the final narrated video using MoviePy and FFmpeg. It includes both a modern React web interface and a standalone command-line tool.

Features

• AI-powered script generation using Google Gemini vision models
• High-quality text-to-speech synthesis via Coqui TTS (offline, no API fees)
• Drag-and-drop web interface built with React, TypeScript, and Tailwind CSS
• Real-time progress tracking with detailed status updates
• Script editing and regeneration without full reprocessing
• Multi-job management with persistent state
• Cross-platform slide conversion using LibreOffice headless mode
• Multiple video codec fallbacks (H.264, MP4V) for broad compatibility
• CLI support via the standalone auto_presenter.py script

Prerequisites

• Python 3.11 or higher
• Node.js 18 or higher
• LibreOffice (headless mode for PPTX-to-PDF conversion)
• FFmpeg (video encoding)
• A Google Gemini API key

Getting Started

Installation

GitHub Codespaces (Recommended):

1. Create a new Codespace from this repository. The devcontainer will automatically install all dependencies.
2. Set up your Gemini API key:

   echo "GEMINI_API_KEY=your_api_key_here" > .env

3. Run the startup script:

   bash start-dev.sh

Local Development:

1. Clone the repository:

   git clone https://github.com/danielcregg/powerpoint-to-video-old.git
   cd powerpoint-to-video-old

2. Install backend dependencies:

   cd backend
   pip install -r requirements.txt
   echo "GEMINI_API_KEY=your_api_key_here" > ../.env

3. Install frontend dependencies:

   cd ../frontend
   npm install

Usage

Web Interface:

1. Start the backend:

   cd backend && python app.py

2. Start the frontend (in a new terminal):

   cd frontend && npm run dev

3. Open http://localhost:3000 in your browser and upload a .pptx file.

Command-Line Interface:

python auto_presenter.py presentation.pptx

API Endpoints:

Method	Endpoint	Description
POST	/upload	Upload a PowerPoint file and start conversion
GET	/status/{job_id}	Get conversion progress and status
GET	/download/{job_id}	Download completed video
GET	/scripts/{job_id}	Get generated scripts for editing
PUT	/scripts/{job_id}	Update scripts and regenerate audio
GET	/jobs	List all conversion jobs
GET	/health	Check service availability

Tech Stack

• Python -- Backend logic and AI orchestration
• FastAPI -- REST API framework with async support
• React 18 -- Frontend UI with TypeScript
• Tailwind CSS -- Utility-first styling
• Google Gemini -- AI vision model for slide script generation
• Coqui TTS -- Offline text-to-speech synthesis
• MoviePy -- Video assembly from images and audio
• PyMuPDF -- PDF-to-image extraction
• LibreOffice -- Headless PPTX-to-PDF conversion
• FFmpeg -- Video encoding and processing
• Vite -- Frontend build tooling

License

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