Skip to main content
Every dictation is saved locally with its audio. The list lives in the Dictation section of the sidebar.

​
What’s stored

For each dictation:
  • The final transcript (post-processed if applicable).
  • The raw transcript before post-processing.
  • The audio as a WAV file.
  • The timestamp, duration, and which model handled it.
  • The app you were focused on when you started.
Files live in ~/Library/Application Support/ai.hyprcore.desktop/. Audio files can grow—a typical minute is around 1–2 MB.

​
Use the history list

  • Click an entry to expand it and see both the raw and post-processed transcripts.
  • Copy to put the transcript on your clipboard.
  • Replay to listen to the audio.
  • Re-process to rerun a different post-processing template against the original transcript without re-recording.
  • Delete to remove a single entry, or use the bulk select.

​
Retention

Dictation can pile up fast. Set a retention window in Settings → Dictation → History retention:
  • Days, weeks, or months. Hyprcore deletes entries older than the cutoff on app start.
  • Off. Hyprcore keeps everything until you manually delete.
Retention applies to both the transcript and the audio file.

​
Export

Right now, history is local-only. To export:
  • Single entry. Use Copy to grab the text.
  • Bulk export. Open ~/Library/Application Support/ai.hyprcore.desktop/history.db with any SQLite viewer, or copy the WAVs directly from the recordings directory.
A built-in CSV/JSON export is on the roadmap.

​
Privacy

Dictation history never syncs to the cloud, even when you’re signed in. It stays on the machine that captured it.