Easy Dictation

Easy Dictation Privacy Policy

Effective date: April 29, 2026
Legal entity / owner: Hlib Rubanskyi

1. Summary

Easy Dictation is a local dictation app for macOS.

  • Audio recording happens locally on the device.
  • Transcription happens locally on the device through the bundled whisper.cpp runtime.
  • The app does not upload audio recordings or transcript text for transcription.
  • The app can download transcription model files from the configured model host when you choose to install a model.
  • The app can place transcript text onto the macOS clipboard and, when you grant automatic paste permission, can send Command-V into the currently focused app.

2. Information the App Handles

  • Microphone audio that you intentionally record while dictation is active.
  • Transcript text generated locally from recorded audio.
  • Local settings such as the selected model, hotkey, and launch-at-login preference.
  • Download state and model files stored locally on your Mac.
  • Clipboard contents temporarily, only to support transcript insertion and clipboard restoration behavior.

3. What Stays Local

  • Microphone audio is recorded locally.
  • Transcription is performed locally.
  • Transcript text is generated locally.
  • Settings are stored locally on the device.
  • The app does not use account creation, sign-in, analytics SDKs, advertising SDKs, or cloud transcription.

4. Model Downloads

When you choose to install a model, the app downloads model files over HTTPS from the configured source.

In the seeded catalog shipped with this build, the model source is Hugging Face - ggerganov/whisper.cpp.

When the app requests a model download, the hosting provider may receive standard network request metadata such as your IP address, user agent, and request timing information in order to serve the file. That network handling is governed by the hosting provider's own terms and privacy practices.

5. Permissions

  • Microphone access, to record dictation audio.
  • Accessibility / automatic paste access, to send Command-V into the focused app after transcription if you enable that workflow.

If automatic paste permission is unavailable, the app can still copy the transcript to the clipboard for manual paste.

6. Local Storage and Retention

  • Installed model files in the app's Application Support directory.
  • Temporary download files while a model download is in progress.
  • User settings in local preferences.

Recorded audio files are temporary working files used during dictation and are removed by the app after the transcription flow completes or is discarded, subject to normal operating-system behavior.

7. Data Sharing

The app does not share microphone audio or transcript text with the developer or with third-party transcription services.

Model downloads are requested from the configured download host, which is a third-party network service for that specific action.

8. Your Choices

  • Decline microphone permission, in which case dictation recording will not work.
  • Decline automatic paste permission, in which case you can still paste transcript text manually from the clipboard.
  • Remove installed models from local storage from within the app.
  • Delete the app and its local support files from your Mac.

9. Children's Privacy

The app is not directed to children under 13. The app does not knowingly collect personal information from children.

10. Changes to This Policy

If the app's data handling changes, this Privacy Policy will be updated before the changed version is released.