Skip to content
WispenRelease 2026.10.05

Version History ​

2026.10.05 ​

  • Added a one-time notice when the Terms of Service or Privacy Policy change.
  • Added Terms of Service and Privacy Policy links to Settings › Help & Feedback.

2026.10.04 ​

  • Redesigned Settings: it opens in the main window like a note, with large buttons, clear pages and short descriptions.
  • The home screen and Mac sidebar now show just Favorites, Uncategorized and your projects. Insights moved to the top of Settings.
  • Recording starts the moment you tap Record, and your first words are no longer cut off.
  • Transcript Cleanup with S1-mini loads while you speak, so stopping is quicker.
  • Speech models now show plain facts (languages, engine, size) instead of speed and accuracy scores.
  • The microphone turns off as soon as you stop, even while Transcript Cleanup finishes.
  • Transcript Cleanup applies to dictations in the Wispen app; Wispen Keyboard dictations use the built-in cleanup.
  • Fixed the stop button sometimes spinning for a long time after a dictation.
  • Fixed other apps' audio sometimes staying quiet after a very short recording (iPhone and iPad).
  • Fixed Settings not opening from the menu bar or with ⌘, when the Wispen window was closed (Mac).
  • Removed Meeting Recorder and Recent Recordings. Recordings kept on the device are deleted.
  • Removed AI Transformations, Quick Reformat and Summarize & Tag from notes, including the summary line at the top of notes.
  • Removed AI Styles from Settings on iPhone and iPad, and settings that didn't do anything: Workspaces, Display Name, Pacing Style and the privacy badges.

2026.10.03 ​

  • Added Apple's SpeechAnalyzer engine, now the default speech model.
  • Added NVIDIA Parakeet speech models: Parakeet v2 for English and Parakeet Ultra for 25 European languages. Parakeet uses your dictionary.
  • Added Transcript Cleanup: download S1-mini by Superwhisper (462 MB) to clean up English dictation on your device.
  • Added push-to-talk on Mac: hold the dictation shortcut, or a single hold-to-talk key such as Fn or Right Option. Double-tap it for hands-free.
  • Added Command Mode on Mac: select text in any app, press ⌥⇧E and say how to change it, using Apple Intelligence.
  • Added AI styles that rewrite your dictation before it's inserted, custom AI modes, and per-app modes on Mac.
  • Added Paste Last Dictation (⌥⇧V), start and stop sounds, and lowering or muting other audio while dictating on Mac.
  • Added word replacements, voice snippets and mid-sentence corrections to the Dictionary. Wispen also offers to add names you correct.
  • Added context-aware recognition: names around your cursor are spelled correctly, on your device.
  • Added the Wispen Keyboard typing layout, which also works without Full Access.
  • Added Read Aloud for notes, with Apple's voice or a natural voice.
  • Added speaker labels for imported meetings and interviews, and Identify Speakers for notes with a recording.
  • Added Insights: words dictated, speaking pace, time saved and streaks.
  • Added Whisper Mode for dictating quietly.
  • Added Shortcuts actions: New Voice Note, Get Last Dictation and Transcribe Audio File.
  • Added ⌘N for a new note and ⌘F to search notes.
  • Added Meeting Recorder, Recent Recordings and Quick Reformat for notes (removed in 2026.10.04).
  • Added a What's New tour, new setup steps for shortcuts and choosing a speech model, and a Licenses page.
  • Esc now cancels dictation into other apps on Mac, and your clipboard is restored after Wispen pastes.
  • Dictation into other apps now inserts the polished text, unless it's typed live as you speak.
  • Subtitle exports and synced playback for imported audio now use the real timing of each sentence.
  • Unused speech models are now released from memory when you switch models.
  • Fixed the Enable Global Shortcuts setting having no effect.
  • Fixed the Wispen Keyboard not using your custom vocabulary.
  • Fixed the menu bar menu always showing Start Dictation during a dictation.
  • Fixed exporting notes on Mac not saving the file.
  • Clearer message when on-device speech recognition can't start, instead of a stuck recording.

2026.10.01 ​

  • Added iCloud sync for notes and projects, with a sync status row in Settings.
  • Added note History: earlier versions are saved automatically and can be previewed and restored.
  • Added Recently Deleted: deleted notes can be recovered for 30 days.
  • Added Undo and Redo in notes.
  • Added Favorites.
  • Added a Dictation Language setting with 10 languages plus automatic detection.
  • Added a Spoken Punctuation setting.
  • Added a setup guide for permissions and the Wispen Keyboard.
  • Added a setting to choose whether Wispen Keyboard dictations are saved as notes.
  • Added this Version History to Settings.
  • Wispen Keyboard dictation now works through the Wispen app when iOS doesn't give keyboards the microphone, stays ready for quick follow-up dictations, and can use your Whisper model.
  • Dictation keeps recording when you lock your phone or switch apps.
  • Dictation now goes where your cursor is in a note, or replaces the selected text.
  • Formatting voice commands now need "Hey Wispen" first, so everyday words are left alone.
  • The microphone permission is now requested when you first need it.
  • The home screen microphone now starts dictating right away.
  • Wispen and the Wispen Keyboard now require iOS 27.
  • Many fixes to dictation, text cleanup, the keyboard, exports and imported audio, including PDF export on iPhone and dictations lost when auto-stop ended a recording or the model changed mid-recording.

2026.09.01 ​

  • Initial release.

Odd Omens is a one-person studio building privacy-first software.