Transcribes audio.mp3 with the base Whisper model, auto-detecting the language and printing a timestamped transcript while writing .txt, .vtt, .srt, .tsv, and .json files next to the audio. Model sizes tiny, base, small, medium, large-v3 trade speed against accuracy. The first run downloads the model weights automatically.
Looking for more? Search all 7,657 commands — works offline, in English or Spanish, and fixes typos.