Significant improvements to OpenAI Whisper and Speaker Diarization speed, accuracy, and model options (including Parakeet streaming, local LLMs like Qwen/Gemma, and Pyannote integration). Added real-time transcription, translation, and in-meeting AI Chat support for live meetings, recordings, and background execution. Redesigned the Live Caption overlay with custom fonts, styles, translucent view modes, and selective translation displays. Introduced a new Dictation mode featuring AI-powered text polishing. Added app video recording capabilities, fixed window capture targeting issues, and enabled automatic background saving. Integrated Mac Calendar integration to automatically name recordings based on event names, with customizable calendar selection. Enhanced export options, including single-file Markdown combining summaries and transcripts, bulk export timestamp toggles, and customized CSV timestamp formats. Streamlined the onboarding setup by integrating core system permission requirements directly into the initial launch process. Continuous general bug fixes, performance improvements, and UX optimizations.

This update brings powerful AI upgrades, smarter automation, and more control over your workflow: • Added support for Qwen ASR V3 and Qwen Aligner models for improved transcription accuracy • Automatic recording for Slack meetings — never miss a conversation • Bulk restart for all failed tasks to save time • New grid layout for a cleaner, more efficient job list • Option to clear cached WAV files and free up space • Customizable export location for transcripts and summaries • Option to include timestamps in exported file names • Recordings now saved in M4A format for smaller file sizes Plus general bug fixes, performance improvements, and overall UX enhancements.

• Voice Profiles for Speaker Identification — Assign real names to speakers using predefined voice profiles • Automatic Speaker Diarisation — Option to auto-detect and separate speakers for recordings and batch jobs • Speaker Voice Preview — Automatically extract and preview a sample voice for each detected speaker • Automatic Speaker Colours — New speakers are assigned distinct colours for easier reading • Improved Transcript Formatting — Group speaker-labeled segments into clean, readable paragraphs • Fullscreen Video Playback with Subtitles — Watch videos fullscreen with perfectly synced captions • Double-Click Segment Playback — Instantly play audio or video by double-clicking a transcript segment • Hide Dock Icon — Run TranscribeX silently in the background without showing a Dock icon • Rename Recordings in Table View — Rename recordings directly without opening details • Selectable Job Descriptions — Easily copy and reference transcription job descriptions • On-Device Transcription — Support for Apple Foundation Models for local transcription • Apple Intelligence Optimisation — Reduced token usage and improved efficiency • Apple Intelligence Summaries — Generate transcript summaries using Apple Foundation Models • Distilled Whisper Models — Faster, more efficient transcription models • New Distilled Language Models — Added German, Korean, and Chinese Whisper V3 variants • Bulk Transcript Export — Export multiple transcripts at once • Folder Monitoring — Automatically transcribe and export files from watched folders • Jobs Continue During Sleep — Transcription tasks keep running while your Mac sleeps • Improved Memory Efficiency — Reduced memory usage and resolved freeze issues • General Performance & Stability Improvements

- Enhanced performance for long videos - Transcribe multiple files at once - Drag-and-drop support for files - Reflow segments by a custom number of characters or words per line - Edit segment time ranges manually - Manually create new segments - Reveal the original audio or video file in Finder - Delete the original audio or video file - Optimized user interface for a smoother experience - Fixed various bugs and improved stability