Transcribe
A voice memo, lecture, interview or video into text and subtitles
Everything runs in your browser, with OpenAI's Whisper model. Your recording is never uploaded. The first time, it downloads the model (about 77 MB), then keeps it.
Drop a recording here or
MP3, M4A, WAV, OGG, FLAC, or an MP4/WebM video
Subtitles (.srt, .vtt) work in VLC, YouTube Studio and most video editors. Check names: Whisper is good, not perfect.