v2.0.2
Latest release
September 2026 · Build 20260903.190244
A complete local transcription workspace
- New desktop interface — responsive light and dark workspaces with full-width page layouts.
- Branded audio recorder with live input level, elapsed time, and microphone selection.
- Stage-aware processing: Ready, Running, and Transcript appear only when relevant.
- Live queue and checkpoints — sequential batch processing with pause, cancel, and resume.
- Transcript workspace: waveform playback, timestamped and full-text editing, find and replace, undo/redo, uncertainty review.
- Non-destructive versions — Original, Cleaned, English, and Bilingual.
- Model manager from Tiny to Large-v3 with download state, hardware guidance, and disk usage.
- In-app history and logs with 20, 50, or 100 items per page.
- Expanded exports — TXT, SRT, VTT, JSON, DOCX, and PDF with branding and watermarking.
Reliability and packaging
- Hardened worker IPC so native engine diagnostics cannot corrupt transcription messages.
- Bundled the matching Python OpenSSL runtime to prevent frozen-app DLL failures.
- Added a unique build number to Settings › About and diagnostic reports.
- Ships as a single per-user Windows installer with a Start-menu shortcut.