Write and edit text
Dictate an email, a document, or a chat message. Wheelhouse puts your words into the text field you’re working in. You say the punctuation, too.
Writing and editing by voiceControl your Windows PC with your voice
When you speak, Wheelhouse transcribes what you're saying and types it into Windows. Some of the things that you say are interpreted as commands that do a wide range of tasks on your computer. You can use your voice for a few tasks or make it the main way you interact with your computer.
Windows 10 & 11 · Free · No account for the default setup
COMMAND EXAMPLE
You say
“hello world period”
A blank document
Hello world.
Your words appear, with the punctuation you asked for.
Try a command below. See what changes.
An interactive illustration, not a recording. This page doesn’t use your microphone.
What you can do
You can start with dictation and add other commands as you need them.
Click buttons and links
Say the name of a button or link. If the name isn’t useful, ask Wheelhouse to show numbers and choose the number you want.
Clicking commandsOpen and switch programs
Open programs, switch windows, and use commands for things you’d normally do with the keyboard. There’s a numbered mouse grid for pointing, too.
Getting around by voiceThere’s also optional AI help for correcting and rewriting selected text. It needs its own setup; dictation and voice commands work without it. Read about text correction.
About Wheelhouse
I want Wheelhouse to be useful when using a keyboard or mouse is painful, difficult, or impossible. Accessibility and reliability are priorities for me.
You can also use it if you prefer talking or want to use your hands less.
I use Wheelhouse myself, but I’m its primary author and it’s only been tested on a limited range of computers. You may run into problems I haven’t seen. If you do, I’d like to hear about them.
Get in touchDavid HiteCreator of Wheelhouse
Questions
Check your computer and get startedDoes my voice leave my computer?
With the default speech recognition, no. Your speech is turned into text on your computer. Wheelhouse sends no telemetry, and dictated text stays out of its logs unless you turn transcript logging on.
Cloud speech recognition is an option you can choose; it sends audio to Google. Optional AI rewriting can also send selected text to the provider you configure. Read about the speech options.
Will it run on my PC?
You need 64-bit Windows 10 or 11, at least 8 GB of RAM, 10 GB of free disk space, and a microphone to use voice control. I recommend 16 GB of RAM and four or more CPU cores. Setup checks your computer before it installs. See the requirements.
Do I need an account or the internet?
You need the internet to install Wheelhouse and download its speech model. The default speech recognition works offline afterward, without an account. Optional cloud features and the hosted help assistant need the internet; their account requirements are separate.
Wheelhouse also asks GitHub once a day whether a newer release exists. The request carries no personal data, but GitHub sees your IP address. To turn the update check off, set CHECK_FOR_UPDATES = false in config.toml.
Is it really free?
Wheelhouse is free and open source. There’s no Wheelhouse subscription. The default local speech recognition is free too. If you choose a cloud service for speech recognition or AI rewriting, that provider’s pricing applies.
What if I get stuck?
You can read the guides, ask the Wheelhouse Assistant, or email for help. The assistant asks you to sign in with a Google Account; the guides and email don’t. Get help.
Install Wheelhouse
Installation takes about 10–20 minutes. The getting-started guide includes a short check in Notepad.