Fact · Dictation on Windows
The Handy README says you can choose Whisper models (Small, Medium, Turbo or Large), with GPU acceleration when a GPU is available.
Checked against its source on October 16, 2026
In the source's own words
“Whisper models (Small/Medium/Turbo/Large) with GPU acceleration when available”
cjpais/Handy: a free, open source speech-to-text app, CJ Pais, GitHub, 2026, Features list, speech models.
Open the original- What it measures
- Handy's README lists its speech models together. The Whisper choices sit next to a Parakeet V3 option that the README describes as CPU-optimized.
- What it does not show
- The README does not say which GPUs are supported on Windows, and it gives no speed figures for the GPU path.