CMPIF2100 Lab Transcriber 2.0

Portable Windows app. Download the .exe below, put it anywhere, and double-click. No Python needed.

The first launch downloads what it needs: the speech model (about 1.5 GB) and, on machines with an NVIDIA card, the GPU acceleration libraries (about 1.3 GB). This happens once; later launches open straight to the main window. You can reclaim that space anytime from Menu > Remove Dependencies.

Records your system audio (Zoom, Teams, recorded lectures, anything through your speakers), transcribes it live on screen, and on stop saves both the recording (.wav) and a clean transcript (.txt). See the README for the full guide.