Minor bug fixes and stability improvements.

This update brings minor fixes and other minor improvements to make VoxScript even better

VoxScript 3.0 brings more ways to record, transcribe and work with your audio—all with on-device processing. • More models: Whisper, Parakeet V3, SenseVoice, Canary and Apple SpeechAnalyzer for supported imports, with Automatic selection by language. • More live-caption choices: English Parakeet streaming and eight-language Nemotron, alongside Whisper's periodic captions. Choose the final transcript model separately. • Share audio or video into VoxScript, confirm your settings and transcribe. Import multiple files with queued processing. • Speaker controls where you need them: expected speaker count, renaming and re-analysis. • More useful transcripts: custom vocabulary, automatic titles, find and replace, reversible filler cleanup and improved editing. • Flexible playback with speed controls and optional quiet-gap skipping. • Export TXT, SRT, VTT, Markdown, JSON, PDF, Word (.docx) and saved audio. • On-device translation with preserved originals, including Canary's supported English-language pairs. • Siri and Shortcuts, recording widgets, Control Center controls and Live Activities with Dynamic Island support. • Recording auto-stop, clear finishing progress, interrupted-work controls and optional App Lock. Model and language availability vary by device and system version. Some models require a first-use download.

Better speaker labels. More dependable recordings. - Improved speaker identification for short replies and overlapping voices. - Audio is saved as you record, with recovery options for interrupted work. - More reliable pause and resume handling when recording is interrupted. - Interrupted transcriptions restart from saved audio to preserve the full recording's context. - Improved Auto language handling for original-language transcription, including Italian and Spanish. - Transcript edits stay consistent in text and subtitle exports. - Retry missing speaker labels from saved audio without using additional transcription minutes. - Improved model setup, file-import error handling, and saved-transcript usage accounting. Thank you for using VoxScript.

Support for more audio and video formats Long recordings transcribe faster, with live progress and time remaining - Interrupted transcriptions now pick up right where they left off - Your finished transcript opens automatically when it's ready - Improved speaker identification - Free plan now includes 30 minutes of transcription - Bug fixes and performance improvements

SPEAKER IDENTIFICATION, SERIOUSLY UPGRADED A new analysis engine identifies who's speaking with far greater accuracy, even in recordings with quick back-and-forth conversation. Each recording now gets its own clean set of speakers, and transcripts split precisely where one person stops and another begins. NAME YOUR SPEAKERS Tap any speaker label to rename it. Turn "Speaker 1" into "Joe" and the name updates across the entire transcript, your history, and every export. SHARING FIXED AND IMPROVED Resolved an issue where the share screen could appear blank. Text, PDF, Word, and SRT subtitle exports, plus Copy to Clipboard, now include speaker labels so you get a properly attributed transcript everywhere. BETTER IN EVERY LANGUAGE Transcribing Japanese, Italian, French, or any of 25+ languages? VoxScript now automatically uses the best model for your language. No settings to fiddle with. FASTER AND SMOOTHER Live transcription stays fast even in hour-long recordings, imported files start processing sooner, and time estimates are now honest. No more lingering "almost done."

Bug fixes and performance improvements.

2.5.0 — Bookmarks, Tags & New Look BOOKMARKS - Pin important moments during recording with one tap - Add bookmarks to any segment after recording - Markers appear on the audio timeline. Tap to jump. - Share 30-second audio clips around any bookmark TAGS - Color-coded labels for organizing transcripts - Predefined tags: Meeting, Lecture, Interview, Personal, Ideas, Work - Create custom tags - Filter by tag in History SEARCH - Results count shown instantly - Search now includes tags - Recent searches remembered AUDIO CLIPS - Long-press any segment to share that audio - Share quotes from interviews or meetings TAPPABLE ENTITIES - Phone numbers, emails, URLs, and addresses detected automatically - Tap to call, email, open, or get directions EXPORTS - Bookmarks and tags now included in all formats RECORDING - Completion screen shows transcription preview, duration, and word count MAC - Keyboard shortcuts: ⌘N new recording, ⌘1/2/3 switch tabs - Drag and drop audio files to import - Native text selection in transcripts - Toolbar search in History NEW LOOK - Teal color scheme - New waveform app icon - Updated typography

• Added faster live transcription with lower latency and improved speech detection • Improved speaker diarization and segment labeling • Enhanced translation and language selection flow • Better export reliability across PDF, SRT, Word, HTML, and text • Updated iPhone and iPad UI polish across key screens • General performance improvements and bug fixes

LIVE RECORDING ENHANCEMENTS • Significantly reduced latency for real-time transcription • Speech-activity detection minimizes silent gaps • Auto-scrolling transcript with quick copy functionality • "Open in History" action after recording completion • Improved microphone handling and interruption recovery UI/UX IMPROVEMENTS • Completely rewritten in Swift for blazing-fast performance • Modern SwiftUI interface with smoother animations • Better word-by-word highlighting during playback • Enhanced model download progress indicators • Clear distinction between Live Recording and File Import modes TECHNICAL IMPROVEMENTS • Optimized background transcription with iOS task management • More reliable notification system for completed transcriptions • Improved memory management for large files • Better handling of special tokens and formatting BUG FIXES • Fixed subscription restoration issues • Corrected audio playback sync timing • Fixed export formatting inconsistencies • Improved stability during long recordings

Faster startup — Transcription models now ship inside the app, so first-time use begins instantly with no downloads required. Improved performance — Speaker identification now runs in parallel with transcription, significantly reducing total processing time.

Updated the app’s subscription system to RevenueCat for more reliable purchases and smoother restore-purchase behavior.

Updated screenshots for clarity.

• Added Privacy Policy and Terms of Use links • Improved subscription information display • Bug fixes and performance improvements

Bug fixes and improvements

A complete rewrite of VoxScript in pure Swift & SwiftUI, featuring: WhisperKit on-device transcription engine Live transcription during recording Apple Intelligence Summaries (iOS 18+) Background transcription with notifications File imports from Files & Photos Word-by-word playback highlighting SwiftData local storage Modern UI with dark theme StoreKit 2 upgrade for purchases Diarization Firebase Analytics (privacy-respecting) Major speed & accuracy improvements Full localization in 8 languages Bug fixes, UI refinements, and smoother exports

NEW FEATURES: • Multi-Language Support - Transcribe in 20 languages including Spanish, French, German, Japanese, Chinese, Arabic, and more • PDF Export - Professional formatted documents with timestamps • Word Export - Open in Microsoft Word or any word processor • Live Recording - Record and see transcription in real-time as you speak • AI Summaries - Generate key points and topics, 100% on-device • Audio Playback - Play recordings with synced transcript highlighting • Edit Transcriptions - Fix any errors after transcription EXPORT OPTIONS: • PDF with formatted timestamps and segments • Word-compatible documents • SRT subtitle files for video editing • Plain text with timestamps IMPROVEMENTS: • Enhanced transcript highlighting synced to audio playback • Improved AI summaries with TextRank algorithm • Tap any timestamp to jump to that point in the audio • Favorite and organize your transcriptions • Faster transcription with optimized Whisper models Privacy-first: Everything happens on your device. No cloud. No accounts. No data collection. Your transcriptions never leave your iPhone.

This update improves the transcription experience and introduces our new pricing model. • You now see more detailed,information while your audio or video is being transcribed. • The app now uses in-app purchases instead of an upfront cost. You receive five free transcriptions, and unlimited transcriptions can be purchased for $2.