Windows
Windows 10 and 11, 64-bit
Signed installer lands with the first public tag.
irm https://sonojot.com/install.ps1 | iex winget
winget install SonoJot.SonoJot Pending acceptance into the community repository.
Press a key. Speak. The finished text lands where you were already typing.
Speech recognition runs on your computer. Windows now, macOS and Linux next.
Cleanup is allowed to tidy. It is not allowed to rewrite.
Remove ‘um’ and ‘uh’, correct supported spellings and vocabulary, and format without rewriting.
So I think we should ship the beta on Friday. Let's keep the old installer around until we're sure.
Whichever you pick, the raw recognition is saved unchanged alongside it.
Your microphone and your computer’s audio, captured together and transcribed.
The words you speak are processed where you speak them.
Builds are published as GitHub releases. The command installs the same build and updates it in place.
Windows 10 and 11, 64-bit
Signed installer lands with the first public tag.
irm https://sonojot.com/install.ps1 | iex winget install SonoJot.SonoJot Pending acceptance into the community repository.
Apple silicon, macOS 14 and later
Microphone capture and clipboard delivery work today. System audio and automatic paste do not.
curl -fsSL https://sonojot.com/install.sh | sh brew install --cask sonojot Pending tap publication.
x86-64, AppImage and .deb
Microphone capture and clipboard delivery work today. System audio and automatic paste do not.
curl -fsSL https://sonojot.com/install.sh | sh curl -fsSLO https://github.com/Cody-Stevens/WindowsSTTApp/releases/latest/download/SonoJot_0.13.0_amd64.deb
sudo apt install ./SonoJot_0.13.0_amd64.deb Available on releases that publish a .deb.
Every release is listed on the releases page. The install commands read that same list, so running one again updates an existing copy and never needs a new URL.