- File Import: Import audio files (WAV, MP3, M4A, FLAC, WMA) and video files (MP4, AVI) for offline transcription
- Drag-and-drop file import support
- Speaker Identification: Identify who said what in a recording — 100% offline
- Downloadable diarization model from Settings
- Configurable expected speaker count per recording
- Color-coded speaker labels in transcript view
- AI Text Enhancement: Grammar correction and tone adjustment powered by Phi-3 (local, offline)
- 6 tone presets: Fix Errors Only, Formal, Professional, Casual, Academic, and Custom
- Side-by-side view: raw transcription vs enhanced text
- Clipboard Auto-Paste: Transcription automatically pasted into your active window
- Accessible Hotkeys: Single-key and sequential hotkey modes — foot pedal compatible
- Large touch targets and enhanced focus indicators (accessibility options)
- Visual Refresh: Modernized UI with improved Settings layout
- Floating recording overlay — draggable, always-on-top, position remembered
- Sound feedback on recording start/stop (configurable)
- Stats dashboard — total words, time saved, recordings count
- Various bug fixes and performance improvements
- Fix: model dropdown not updating after downloading new models in Settings
- Fix: dropdown losing selected model after closing Settings
- Fix: safe fallback when active model is uninstalled (no more hardcoded "base")
- Fix: prevent multiple app instances (file lock)
- Fix: model state not refreshing after downloading in Settings
- Fix: auto-download base Whisper model during publish if missing
- Fix: native Whisper library loading in MSIX sandbox
- Fix: database concurrency crash on second recording
- Fix: onboarding hotkey default not persisted
- Fix: Store asset image dimensions for WACK validation
- Include base Whisper model in installer — no download needed on first use
- Fix: hotkey now updates immediately when changed in Settings or Onboarding
- Fix: allow empty hotkey to disable global hotkey
- Rename "Set" button to "Change Hotkey"
- Fix: Onboarding Step 2 "Next" disabled until a model is installed and active
- Fix: "Active" badge only shown when model is both installed and selected
- Add bug report page on website
- Add releases page on website
- Fix: MSIX bundle for x64 + ARM64 Store submission
- Fix: correct architecture-specific DLL selection
- Add WACK test runner script
- Fix: graceful handling when no Whisper models installed
- Fix: revert to Whisper.net.Runtime + smarter native DLL discovery
- Fix: use Windows-specific Whisper runtime + fix MSIX native DLL layout
- Fix: native DLL discovery for Store package
- Version bump for Store resubmission
- Initial release on Microsoft Store
- Speech-to-text transcription using OpenAI Whisper (100% offline)
- Support for 99 languages with auto-detection
- Toggle-to-Talk and Push-to-Talk recording modes
- Global hotkey support (Ctrl+Shift+R default)
- Auto-copy transcription to clipboard
- Dark/Light/System theme support
- First-run onboarding wizard
- Model management (download Tiny/Base/Small/Medium/Large)
- System tray integration with minimize-to-tray
- Recording overlay window
- Audio playback with progress slider
- Settings persistence (SQLite)
- Update checker
- Privacy-first: no data leaves your machine
