What is AI YouTube Shorts Generator?
AI YouTube Shorts Generator is an open-source tool that automatically transforms long-form videos into engaging vertical shorts optimized for TikTok, YouTube Shorts, and Instagram Reels. It uses GPT-4o-mini for intelligent highlight selection and Whisper for accurate speech-to-text transcription. The tool analyzes your video content, identifies the most engaging segments, and produces polished vertical videos complete with stylized subtitles and smart cropping.Key Features
Flexible Input
Supports both YouTube URLs and local video files for maximum convenience
GPU Acceleration
CUDA-enabled Whisper for fast speech-to-text (transcribe 5min video in ~30s)
AI Highlight Selection
GPT-4o-mini automatically finds the most engaging 2-minute segments
Interactive Approval
Review and approve/regenerate AI selections with 15-second auto-approve timeout
Auto Subtitles
Stylized captions with customizable fonts burned directly into video
Smart Cropping
Face-centered crops for talking heads, motion tracking for screen recordings
Vertical Format
Perfect 9:16 aspect ratio optimized for mobile viewing
Automation Ready
CLI arguments, auto-quality selection, and timeout-based approvals for batch processing
How It Works
The tool processes videos through an 11-step pipeline:AI Analysis
GPT-4o-mini analyzes transcript to select the most engaging 2-minute segment based on criteria like interest, usefulness, surprise, controversy, or thought-provoking content
Interactive Approval
Review the AI’s selection with options to approve, regenerate, or cancel. Auto-approves in 15 seconds if no input provided
Smart Crop
Applies intelligent cropping:
- Face videos: Static face-centered vertical crop (no jerky movement)
- Screen recordings: Half-width display with smooth motion tracking (max 1 shift/second)
Output Format
The tool generates videos named with the pattern:{video-title}_{session-id}_short.mp4
Example: my-awesome-video_a1b2c3d4_short.mp4
- Slugified title: Lowercase with hyphens instead of spaces
- Session ID: 8-character unique identifier for concurrent execution support
- Resolution: Matches source video height (720p → 404x720, 1080p → 607x1080)
The session ID system allows you to run multiple instances simultaneously without file conflicts.
When to Use This Tool
Content Repurposing
Transform existing long-form content (YouTube videos, webinars, podcasts) into bite-sized shorts for social media
Highlight Reels
Automatically extract the most engaging moments from interviews, presentations, or tutorials
Social Media Marketing
Create vertical video content optimized for TikTok, Instagram Reels, and YouTube Shorts
Batch Processing
Process multiple videos in sequence with automation flags for hands-free operation
System Requirements
Recommended:- Python 3.10 or higher
- NVIDIA GPU with CUDA support (for fast transcription)
- FFmpeg with development headers
- ImageMagick (for subtitle rendering)
- 4GB+ RAM
- Python 3.10 or higher
- FFmpeg
- ImageMagick
- 8GB+ RAM (transcription will be significantly slower)
What’s Next?
Quickstart
Generate your first short in under 5 minutes
Installation
Complete setup instructions for your platform