v1.0.0 - Windows demo build

Download & run (no setup required)

Download audio_transcribe-windows-x64.zip below, unzip it (keep all files together), open PowerShell or Command Prompt in the folder, and run:

.audio_transcribe.exe meeting.wav     # transcribe an existing audio file
.audio_transcribe.exe                 # record now (Enter to stop), then transcribe
.audio_transcribe.exe --help          # all options

This is the full record -> transcribe -> summarize pipeline bundled with PyInstaller, so no Python install is needed.

Notes

• Runs on CPU and defaults to the small Whisper model so it works on any Windows machine. First run downloads the model (~0.5 GB) and caches it. For higher accuracy pass --model large-v3. For GPU acceleration, run from source.
• The summary step needs an ANTHROPIC_API_KEY in your shell. Without one it still records + transcribes and skips the summary.
• Recording uses Windows WASAPI (default speakers + default mic).

See the README for full docs and how to build the exe yourself.