Real-time transcription and translation of any app's audio — powered by whisper.cpp on your Mac's Metal GPU. Completely private, fully on-device.
Features
Record any app's audio system-wide — Zoom, YouTube, podcasts — and watch the transcript appear word by word with simultaneous translation flowing alongside it.
Drag & drop any audio or video file. WhisperASR transcribes it with a live progress bar and time estimate. MP3, M4A, MP4, MOV and more.
Every segment shows the original transcript alongside your translation in teal. Export the transcript to a text file when done.
Whisper runs on-device via whisper.cpp. Your audio and transcripts never touch a server.
Leverages Apple's Metal framework for fast transcription on M-series chips.
Translation uses any OpenAI-compatible endpoint — including local models.