Generate subtitles or transcriptions from your media files or video URLs using OpenAI's Whisper model. Supports various video platforms including YouTube, Vimeo, Dailymotion, and more.
- Modern Gradio web interface with two-column layout
- Prominent Help, Tips & Model Guide section
- Custom footer and author credit
- Upload audio/video files or paste a video URL (YouTube, Vimeo, Dailymotion, Twitch, etc.)
- Supported file types: mp3, wav, m4a, mp4, avi, mov, mkv
- Choose Whisper model size and device (CPU, CUDA, Apple Silicon MPS)
- Model download/caching in project folder or system cache
- Remove downloaded models and clear transcription cache with one click
- Output preview, full output, and downloadable SRT or TXT file
- Output file is named after the video title or uploaded file
- Progress bar and status messages for all steps
- Clone this repository
- (Recommended) Create and activate a Python virtual environment
- Install dependencies:
pip install -r requirements.txt
- Run the app:
python app.py
- Open the local URL shown in your terminal (usually http://127.0.0.1:7860)
- Upload a media file or paste a supported video URL
- Adjust settings as needed (model size, device, language, etc.)
- Click Generate
- Download your output file (SRT or TXT)
- Output files are named after the video title (for URLs) or the uploaded file name.
- Models are cached in the project
models/folder or in your system cache (~/.cache/whisper). - Use the "Remove Downloaded Models" button in Advanced Settings to clear model files.
- Use the "Clear Transcription Cache" button to remove cached outputs.
- The app supports non-commercial use only. See LICENSE for details.
- Downloading video/audio from third-party platforms may be subject to copyright laws. You are responsible for ensuring your use complies with all applicable laws and platform terms.
See LICENSE for non-commercial use terms and copyright/copyright-disclosure.
Created by Max Lee