← Back to VoxScribe Guide

Offline Whisper Dictation on Windows, No GPU Needed

Whisper is OpenAI's open speech recognition model. VoxScribe runs it on your own PC, so you can dictate into any Windows app with no internet connection, no account and no audio leaving the machine.

Download VoxScribe (free) View source on GitHub

Free & open source (MIT) · Windows 10/11 · CPU only

How it runs

VoxScribe uses faster-whisper, a fast implementation of Whisper, on the CPU. The speech model downloads once on first run and is stored on your computer. After that you hold F9 in any app, speak, and release, and the transcribed text is typed where your cursor is.

The models

Real speeds on a CPU

These numbers come from a 15 second spoken clip on an older laptop, using the CPU only (its graphics card is too old to help). Your computer will differ, but the ratios hold up.

The larger models are worth it for careful work, but for everyday dictation Small is quick enough that the text appears a few seconds after you let go of the key.

Fast first, careful when you need it

Live dictation uses a fast setting so the text appears quickly. If a sentence comes out wrong, press F8 within two minutes and VoxScribe reruns that clip on a bigger model with a more careful search, then replaces the text in place. You get speed by default and accuracy when it matters, without changing settings.

Frequently asked questions

Do I need a graphics card to run Whisper dictation on Windows?

No. VoxScribe runs Whisper on the CPU, so a graphics card is not needed. On a laptop with an older graphics card that does not help, a 15 second clip took about 4 seconds with the Small model.

Which Whisper model should I pick?

Small is the default and a good balance. Choose Base if your computer is slow and you want faster results, and choose a larger Pro model if accuracy matters more than speed. You can also press F8 after any dictation to rerun that clip on a more careful model.

Is anything uploaded when I dictate?

No. Audio is transcribed on your computer and discarded. The only download is the speech model itself, once, on first run.

Download VoxScribe (free) View source on GitHub

Related guides: Free Voice Dictation Guide, Dictate Into Any App Guide, Offline Alternative to Windows Voice Typing, Arabic Voice Typing for Windows, Free .srt Subtitle Generator, Dictate AI Prompts, Voice Typing for Wrist Pain, Android Voice Typing, Android Voice Typing Without the Cloud, Privacy at a Glance, Windows Security Warning Explained.

Home · Android · GitHub · Privacy · This site uses Google Analytics, see the privacy policy.