What's new since v0.4
- Downloadable model catalog — download speech recognition models directly in the app: Breeze-ASR-25 (default, best for Mandarin/Taiwanese-accented speech) plus the official whisper.cpp multilingual models — Large v3 Turbo (~1.6 GB), Medium (~1.5 GB), Small (~488 MB), Base (~148 MB), and Tiny (~78 MB). Manage everything in Settings → Speech Recognition Models: per-model download with progress and resume, and delete to free disk space.
- Choose your transcription model — a new picker in the toolbar switches between downloaded models; the change takes effect on the next transcription (including re-transcribe and live recording). Smaller models trade accuracy for speed.
- Lighter first run — the initial setup sheet now offers a choice of model, so you can start with a small download instead of the 3 GB default.
Install
Download WhisperASR.zip, unzip, and move WhisperASR.app to /Applications. Signed with Developer ID and notarized by Apple. Requires macOS 14+ (Apple Silicon).