Handy

Press a shortcut, speak, and the words appear in whatever app you are using, transcribed on your own computer.

Instead of Otter.ai, cloud voice typing

github.com/cjpais/handy

Screenshot from github.com/cjpais/handy, taken 2026-10-30, shown to help you recognise the tool. It belongs to its maker. Ask us to remove it.

Anyone who has dictated a message to a phone and watched “meet at six” come back as something unprintable knows that speech to text is mostly a question of whose computer is listening. Handy runs the listening on yours. You press a shortcut, speak, and the text is pasted into whatever app is in front of you, using Whisper or Parakeet V3 models that run locally. The project calls itself completely offline and says your voice stays on your computer.

Handy scores a 5 on lock-in, a 4 on footprint, self-direction, power and delivery and a 3 on longevity.

One developer makes it, with four sponsors listed on the project page. The code is MIT licensed, but the Handy name and logo are not open source, so forks have to use their own branding. Models download over the network the first time, and the project offers a manual install for locked-down networks. We have not tested accuracy, so Delivery rests on what the project says it does.

What stands out

• Dictates into any app, with the transcription done on your own computer
• MIT licensed, free, and made by one developer with four listed sponsors
• The Handy name and logo are not open source, only the code

For comparison, we also reviewed Otter.ai (14/30) as a benchmark.