Skip to content

FatFatYoung/HtmlToVideoPixie

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Html To Video Pixie v1.0.0

Turn text prompts into videos via AI-generated HTML. No coding required.

Html To Video Pixie Logo

FeaturesQuick StartHow It WorksComparisonDownload


📖 Overview

Html To Video Pixie is an AI-powered desktop application that transforms natural language descriptions into HTML/CSS animations, then renders them as high-quality MP4 videos or GIF animations.

Simply describe what you want in plain language, and the AI generates the HTML code. The built-in rendering engine converts it to video — no coding skills required.

✨ Features

Core Capabilities

  • Natural Language to Video — Describe your animation in words, get a video
  • Real-time Preview — See the HTML animation instantly as AI generates it
  • Dual Output Formats — Export as MP4 video or animated GIF
  • Built-in Code Editor — Fine-tune the generated HTML directly in the app

AI Integration

  • Multi-Model Support — OpenAI, DeepSeek, Qwen, Zhipu AI, Xiaomi MiMo, and more
  • Custom Providers — Add any OpenAI-compatible API service
  • Streaming Output — Watch AI generate code in real-time

User Experience

  • Bilingual Interface — English and Chinese with one-click switching
  • No Installation Hassle — Pre-packaged installer with FFmpeg included
  • Auto-save — All creations and chat history preserved automatically
  • Import HTML — Load existing HTML files for rendering

🚀 Quick Start

Option 1: Download Installer (Recommended)

  1. Download HtmlToVideoPixieInstaller.exe
  2. Run installer (requires administrator privileges)
  3. Launch from desktop shortcut or Start Menu

Option 2: Run from Source

# Install dependencies
pip install playwright requests Pillow tkinterweb

# Install Playwright browsers
playwright install chromium

# Ensure FFmpeg is in PATH
ffmpeg -version

# Run the application
python main_tkinter.py

🎯 How It Works

┌─────────────────────────────────────────────────────────────────┐
│                    Html To Video Pixie                          │
├─────────────────────────────────────────────────────────────────┤
│                                                                 │
│   ┌──────────────┐     ┌──────────────┐     ┌──────────────┐    │
│   │   User Input │ ──► │   AI Model   │ ──► │  HTML/CSS    │    │
│   │  (Natural    │     │  (GPT-4o,    │     │  Animation   │    │
│   │   Language)  │     │  DeepSeek..) │     │   Code       │    │
│   └──────────────┘     └──────────────┘     └──────┬───────┘    │
│                                                    │            │
│                                                    ▼            │
│   ┌──────────────┐     ┌──────────────┐     ┌──────────────┐    │
│   │   MP4/GIF    │ ◄── │    FFmpeg    │ ◄── │  Playwright  │    │
│   │   Output     │     │   Encoding   │     │   Browser    │    │
│   └──────────────┘     └──────────────┘     │  Rendering   │    │
│                                             └──────────────┘    │
└─────────────────────────────────────────────────────────────────┘

Technical Stack:

  • GUI: Python tkinter (lightweight, cross-platform)
  • AI: OpenAI-compatible API with streaming
  • Rendering: Playwright (headless Chromium/Edge)
  • Encoding: FFmpeg (MP4/GIF output)

🆚 Comparison

vs Other HTML-to-Video Tools

Project Core Idea Target Users Tech Stack
HtmlToVideoPixie Natural Language → HTML → Video Non-programmers, creators Python, Tkinter, AI API
Remotion React Components → Video Frontend developers React, TypeScript, Node.js
HyperFrames HTML Source → Video AI Agents, developers HTML, CSS, JavaScript
html-video Data → Multi-frame HTML → Video Automation, Coding Agents Python, Chrome, FFmpeg

vs AI Video Models (Sora/Kling)

Aspect HtmlToVideoPixie Sora/Kling
Control Precise, deterministic (code-based) Probabilistic (diffusion-based)
Text Rendering Vector-sharp, perfect for subtitles Often garbled/unreadable
Resolution Unlimited (4K/8K possible) Limited to 1080p
Style Geometric, data-viz, presentations Photorealistic, cinematic
Editability Direct code editing Re-prompt required

Unique Advantages

  • Exact Timing Control — "Rotate 45° at exactly 3 seconds" is possible
  • Crystal Clear Text — Perfect for data visualization, PPT animations, subtitles
  • Infinite Scalability — Vector graphics scale to any resolution
  • Code-Level Precision — Every pixel is deterministic and reproducible

📁 Project Structure

HtmlToVideoPixie/
├── main_tkinter.py          # Main application
├── installer.py             # Installer script
├── build_installer.bat      # Build script
├── logo.ico                 # Application icon
├── logo.png                 # Original logo
├── config/
│   └── providers.json       # AI provider configuration
├── core/
│   ├── ai_client.py         # AI API client
│   ├── video_generator.py   # Video rendering engine
│   └── i18n.py              # Internationalization
├── README.md                # This file
└── LICENSE                  # MIT License

⚙️ Configuration

Adding Custom AI Providers

  1. Open Settings → Model Settings
  2. Click Add Provider
  3. Fill in:
    • Name: Display name (e.g., "My DeepSeek")
    • API URL: Base URL (e.g., https://api.deepseek.com/v1)
    • API Key: Your API key
    • Model ID: Model name (e.g., deepseek-chat)
  4. Click Save

Configuration is stored in %APPDATA%\HtmlToVideoPixie\providers.json

Supported AI Services

  • OpenAI (GPT-4o, GPT-4-turbo)
  • DeepSeek (deepseek-chat, deepseek-coder)
  • Qwen (qwen-turbo, qwen-plus)
  • Zhipu AI (glm-4)
  • Xiaomi MiMo
  • Any OpenAI-compatible API

💡 Tips for Best Results

Prompt Engineering

Good Prompt:

Create a colorful bar chart animation showing monthly sales data:
- 5 bars for Jan to May
- Each bar grows from bottom with a smooth ease-out animation
- Bars appear one by one with 0.3s delay
- Show value labels on top of each bar
- Use gradient colors: blue to purple

Why it works:

  • Specific element count and names
  • Clear animation description
  • Timing parameters
  • Visual style details

Editing Workflow

  1. Generate — Let AI create initial code
  2. Preview — Check animation in preview pane
  3. Edit — Adjust values in the code editor
  4. Iterate — Regenerate if needed
  5. Export — Save as MP4 or GIF

📋 System Requirements

  • OS: Windows 10/11 (64-bit)
  • RAM: 4GB minimum, 8GB recommended
  • Disk: 500MB free space
  • Network: Required for AI API calls

🔧 Troubleshooting

Issue Solution
FFmpeg not found Install FFmpeg and add to PATH
Browser launch failed Run playwright install chromium
API connection error Check API key and network
Video generation slow Reduce duration/FPS in settings

📄 License

MIT License - See LICENSE for details.

🙏 Acknowledgments

📧 Contact


Made with ❤️ by FatFatYoung

About

Turn text prompts into videos via AI-generated HTML. No installation needed.

Topics

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors