Paste the full link to the YouTube video. The system will extract the audio and generate a summary.
The backend for this feature will need to use a library like **yt-dlp** or a cloud service to download the video's audio stream before it can be sent to the Speech-to-Text model.