Release Notes

See what's new in each version of Chirp.

Download v1.3.2

v1.3.2

LatestApril 2026
  • ImprovedMedia pause now instant — persistent audio monitor detects playback with zero delay
  • ImprovedCorrectly detects playing vs paused media — no more accidental playback starts
  • ImprovedWorks with any audio source — Spotify, YouTube, browser media, and more
  • ImprovedModel downloads continue when settings window is closed
  • ImprovedModel sizes in settings now show actual disk usage
  • ImprovedMedia playback state included in diagnostic reports
  • RemovedmacOS notification for meeting transcription completion

v1.3.1

April 2026
  • NewAuto-pause media — playing music or videos pause when recording starts, resume when done
  • NewSystem Audio Recording Only — no longer requires full Screen Recording permission
  • ImprovedNotch messages auto-size to fit text — no more clipped messages
  • ImprovedConcurrent model downloads — solo and meeting models download independently
  • ImprovedCleaner download progress UI with inline layout
  • ImprovedTrial users now have full access to meeting mode
  • FixedAlignment model failed verification after download (zip extraction bug)
  • FixedDownload progress UI glitched when downloading multiple model groups
  • RemovedRedundant macOS notifications for model errors and failed transcriptions

v1.3.0

April 2026
  • NewSpeaker profiles — voices are recognized across meetings automatically
  • NewBatch actions — re-transcribe, tag, or delete multiple recordings at once
  • NewSpeaker library — see all identified voices and single appearances
  • ImprovedBetter speaker detection — fewer phantom speakers from short interjections
  • ImprovedHandles overlapping speech — mixed-voice segments excluded from analysis
  • ImprovedUpgraded voice recognition model for more accurate speaker matching
  • ImprovedDeterministic clustering — same recording always produces the same result

v1.2.0

March 2026
  • ImprovedSmarter speaker detection — automatically finds the right number of speakers
  • ImprovedBetter speaker accuracy — fewer false speaker switches, especially with 3+ speakers
  • ImprovedNo recording time limit — meetings can run as long as needed
  • ImprovedMore reliable punctuation on all audio quality levels

v1.1.0

March 2026
  • NewMeeting transcription with automatic speaker labels and voice profiles
  • NewRe-transcribe any recording — choose Parakeet (fast) or Whisper (precise)
  • NewRun meeting diarization on any recording from the right-click menu
  • NewLive processing indicator with current step and elapsed time
  • NewInterrupted meetings resume automatically on next launch
  • NewSpeaker count filter in history (1, 2, 3, 4+)
  • NewCollapsible filter sections
  • NewDelete audio only — keep the transcription, free up space
  • ImprovedHistory cards — cleaner layout, play button grouped with duration
  • ImprovedLicense page — side-by-side feature comparison
  • ImprovedFaster solo transcription — model pre-loads during recording
  • ImprovedBetter memory management — models released when not in use
  • ImprovedWord highlighting stays accurate after scrubbing
  • ImprovedSidebar text no longer truncates on selection
  • FixedWord spacing errors in solo transcription
  • FixedNumbers no longer merge with preceding words
  • FixedContext menu disabled during processing
  • FixedRe-transcription preserves original recording mode

v1.0.0

March 2026
  • New100% local transcription — your voice never leaves your Mac
  • NewUp to 190x realtime speed powered by Apple Neural Engine
  • NewSolo Mode — tap hotkey to dictate, text appears at your cursor
  • NewMeeting Mode — on-device speaker identification
  • NewLives in your MacBook notch — always one tap away
  • NewSearchable history with tags, filters, and audio playback
  • NewWord-level timestamps with playback highlighting
  • NewExport to Markdown, SRT subtitles, or plain text
  • NewCustomizable hotkeys and sound effects