Whisper Speech support

Whisper Speech is a local speech-to-text app for Apple Silicon Macs running macOS 14 or later. It uses whisper.cpp to transcribe recordings and imported audio on your Mac.

Contact

For help, email contact@oddly-interesting.org. To report a bug, email bug@oddly-interesting.org.

Include the app version, macOS version, Mac model, the steps you took, and what happened. Remove private information from any screenshots or examples. The app does not automatically attach recordings, transcripts, or logs to your message.

Getting started

Choose a recognition model and select Download model. The app shows its size before downloading. Internet access is needed for this download; after the model is installed, transcription works offline.

Use the recording button to record speech, then stop to transcribe it. You can edit the result, copy it, or export it as a TXT file. You can use the button and copy text manually without enabling the global hotkey or automatic pasting.

Permissions

Permission controls are in macOS System Settings → Privacy & Security. If automatic pasting is unavailable, use the copy button and paste manually.

Importing files

Import supports MP3, M4A, WAV, MP4, OGG, OGA, and Opus. OGG-family decoding depends on the codecs available in your macOS version; the app reports an error if a codec is unavailable and does not install additional packages.

Models and saved data

Use Models to download, remove, or redownload a recognition model. In Settings, you can manage the history limit and delete saved audio. Recordings and transcripts remain local unless you choose to copy, export, or share them. See the privacy policy for storage, retention, and download details.