Whisper-WebUI

Why we included this project

Making subtitles usually means juggling a transcription tool, a caption editor, and a pile of command-line flags. Whisper-WebUI folds that into a single browser interface: choose the Whisper variant that matches your speed or accuracy needs, then transcribe from an uploaded file, a YouTube link, or a live microphone. Results can be saved as SRT, WebVTT, or plain text. The pipeline gets more interesting before and after transcription: optional Silero VAD trims silence, UVR can strip background music, and pyannote speaker diarization labels who spoke when, which helps a lot with interviews and podcasts. Translation is built in as well, both speech-to-text into English and whole-file translation using NLLB or the DeepL API.