YouTube Transcript, Translation & Summary (Whisper + Hugging Face)

This Space extracts audio from a YouTube video, detects language, transcribes speech using OpenAI Whisper, translates to English if needed, and provides a summary.

How to use:

Paste a YouTube URL in the box.
Click "Process".
View detected language, full transcript, English translation, and summary.

Tech stack: Gradio, Hugging Face Transformers, OpenAI Whisper, Facebook BART, yt-dlp

Notes

Works best with videos under 10-15 minutes.
For non-English videos, automatic translation to English is performed.
For longer videos, only the first ~30 minutes may be processed due to resource limits.

Requirements

All dependencies listed in requirements.txt.
yt-dlp is included as a pip dependency.

Author

Built for quick demo and prototyping by Jagan (template by ChatGPT).