Video Captioner
AI-Powered Transcription & Translation Tool
Generate subtitles from any video with automatic speech recognition. Translate to 15+ languages. All processing happens locally on your computer - no uploads, no file size limits, 100% private.
Download Desktop Tool
Download the GUI app (Python 3.8+). On first run, dependencies install automatically.
Windows: double-click the .pyw file. Mac/Linux: use the .py download below.100% Private
All processing happens locally on your computer. Your videos never leave your device.
15+ Languages
Auto-detect source language and translate to multiple target languages.
YouTube Support
Download and caption videos directly from YouTube URLs.
Burn Captions
Export SRT files or burn captions directly into your video.
Requirements
-
Python 3.8+ required. Dependencies (Whisper, yt-dlp, deep-translator, FFmpeg) install automatically on first run.
How to Use
Windows (.pyw): Double-click the downloaded file
Windows (.py): Right-click → "Open with" → "Python"
Mac/Linux: Terminal →
python3 video_captioner.py
-
Launch the application
Double-click the downloaded file (or use Python for .py files)
-
First-time setup
- Click "Install" when prompted for dependencies
- Wait 5-10 minutes for automatic installation
- App will restart automatically when complete
-
Configure your settings
- Browse for a local video or paste a YouTube URL
- Choose source language (or auto-detect)
- Optionally translate to another language
- Select output format (SRT file or burned captions)
- Choose AI model size (base recommended)
-
Generate captions
Click "Generate Captions" and watch the progress
-
Done!
Your captioned file will be saved to your chosen output folder
Platform Matrix
- Windows: Use Python `.pyw` for double-click launch.
- macOS: Use Python `.py` from Terminal.
- Linux: Use Python `.py` or CLI script.
Troubleshooting
- Install errors: Run again with internet access and admin permissions if package installation fails.
- Burn captions disabled: FFmpeg was not detected; generate SRT output or install FFmpeg manually.
- Slow processing: Use `tiny`/`base` with `fast` profile for draft outputs.