Release Notes
Every update, documented.
v1.5.0Latest
New Features- Translation for more languages — When Apple Translation doesn't support a language pair (for example Hebrew → English), Glasscribe now translates on-device with Firefox Translations by Mozilla. The language model downloads once on first use (about 20–65 MB), and translation runs entirely on your Mac. Pairs without English translate through English (for example Hebrew → Korean).
- Translate & Paste for Dictation — A new toggle next to fn Hold-to-Dictate translates what you dictate into your Target Language before pasting. If translation fails, your original text is pasted instead.
Improvements- The menu bar shows translation model download progress, and Start waits until the model is ready.
v1.4.1
Bug Fixes- Floating window drag on macOS 27 — The transcript window could no longer be moved by dragging after updating to macOS 27. A SwiftUI change in macOS 27 stopped background drags from reaching the window; Glasscribe now handles the drag itself, so you can grab the window anywhere (including over transcript text) and move it as before.
v1.4.0
New Features- Clean up with Apple Intelligence — An optional Dictation setting that removes filler words (um, uh), stutters, and repeated words before pasting. Runs entirely on-device. If the cleanup would change any of your words, Glasscribe pastes your original text instead. Requires Apple Intelligence; currently works best in English.
Improvements- Ready for macOS 27 Golden Gate — the audio pipeline is hardened against a speech framework change in macOS 27.
- Updated the auto-update framework with the latest security fixes.
v1.3.0
Pricing- One-time purchase only — Glasscribe is now a single $12 lifetime license. Monthly and annual subscription plans have been retired in favor of a simpler, no-strings pricing model.
Improvements- Streamlined Settings — the Subscription tab is now a focused License tab.
- Lighter license validation path and smaller in-app footprint.
v1.2.3
Improvements- Private auto-paste — Auto-pasted transcriptions no longer appear in your clipboard manager's history (Paste, Maccy, Alfred, LaunchBar, PopClip, and others that follow the nspasteboard convention).
- Performance and bug fixes.
v1.2.2
Improvements- Minor bug fixes and stability improvements.
v1.2.1
Improvements- Smarter locale defaults — Language picker now selects the most common region by default (e.g. English → en-US instead of en-AU).
Bug Fixes- Fixed translation download dialog appearing repeatedly when switching between regions of the same language.
v1.2.0
New Features- Dictation engine fallback — 12 new languages now available via Dictation engine: Arabic, Danish, Dutch, Finnish, Hebrew, Malay, Norwegian, Russian, Swedish, Thai, Turkish, and Vietnamese.
- Dynamic language detection — Available languages are detected at runtime. New languages added by Apple appear automatically.
- Smart translation targets — Translation target picker only shows language pairs actually supported by Apple Translation.
- Download progress — See real-time percentage when downloading speech model assets.
Improvements- Per-line translation tracking — Each line remembers its source language. Changing languages only re-translates matching lines instead of everything.
- Faster engine loading — Fixed a deadlock that could prevent speech models from loading after language switches.
- Post-download format fix — Resolved an issue where starting transcription failed for newly downloaded languages.
v1.1.8
Bug Fixes- Minor bug fixes and stability improvements.
v1.1.7
New Feature- Overlay size controls — Adjust overlay width and height from Settings → Appearance.
- Accent color picker — Choose from 6 color presets to personalize the overlay.
Bug Fixes- Fixed overlay scrolling issues.
Improvements- Reduced disk writes during active transcription.
- The app now checks for updates automatically on launch.
v1.1.6
Improvements- What's New popup — See release notes automatically after each update.
- Automatic update checks — The app now checks for updates on launch and every 24 hours, so you never miss a new version.
v1.1.5
New Feature- Dictation Mode — Press the fn key to instantly start dictating. Type with your voice anywhere you need text input.
Performance- Reduced memory usage and system resource consumption for a lighter, more stable experience.
v1.1.4
Improvements- Bug fixes and stability improvements
v1.1.3
Performance- Optimized CPU and memory usage during recording