Hey HN,

I’ve been working on WizWhisp, a Windows desktop app for running OpenAI’s Whisper locally. I wanted something native, simple, and privacy-friendly. Existing tools are great but didn’t quite fit what I needed, so I built my own.

What it does:

Drag in audio/video and get transcripts (TXT, SRT, VTT).

Runs Whisper 100% offline.

Uses CUDA if you have an Nvidia GPU; otherwise falls back to CPU.

Handles long recordings.

Tech: C# + WinUI3 for UI, whisper.cpp for inference.

Business model: Free for standard transcription. A one-time Pro upgrade (no subscription) unlocks batch processing and removes the length limit when using the Large model.

Happy to hear any feedback or feature ideas.