100% Private & Secure
Your media never leaves your device. All transcription happens locally using your CPU or GPU. Perfect for confidential meetings, interviews, and sensitive content.
Latest version is now available
Transcribe Audio & Video
Private, Fast, & Local.
Unlock the power of OpenAI's Whisper model directly on your desktop. No cloud subscriptions, no data leaks, just pure GPU-accelerated speed with secure media playback for transcript review.
2.0.0: Embedded media preview • Clickable transcript segments • Improved subtitles
- Open Source under MIT License
Why Choose WhisperDesk for Audio & Video Transcription?
We built this tool because uploading sensitive meeting recordings to the cloud is risky. WhisperDesk runs entirely on your hardware for complete privacy, from transcription to media review.
Secure Media Preview
Play local audio and video beside your transcript through a private whisperdesk-media:// protocol that only serves approved files.
Clickable Transcript Timeline
Timestamped transcript segments become navigation controls. Click a line to jump directly to that moment in the original recording.
GPU Accelerated Speed
Optimized for Apple Silicon (Metal) and NVIDIA (CUDA) for blazing fast transcription speeds up to 100x real-time.
Multi-Format Export
Export to SRT, VTT, plain text, DOCX, PDF, or Markdown. Improved segmentation keeps subtitles easier to scan and edit.
Batch Processing
Drag and drop multiple files. Process unlimited audio or video files sequentially with automatic queue management, resume, retry, duplicate protection, and live ETA.
Open Source & Community Driven
WhisperDesk is free software licensed under MIT. Transparency is key when handling private data. Check our code, contribute features, or build your own fork. Join our growing community of developers and users.