Hold fn. Speak. Done.
Your words appear live as you talk, get cleaned up when you let go, and paste into any app, all offline on your Mac.
Demo: holding the fn key and saying “um so I think we should uh push the deadline to friday no wait thursday”. Words appear live in a bubble next to the cursor. After letting go, bolipi pastes “So I think we should push the deadline to Thursday.”.
Demo
Said messy. Pasted clean.
A replay of what bolipi does on your Mac. Grey guesses turn blue as they lock in, then the clean version lands in the app. Pick a sentence and press play, or hold the key to step through it.
Just a demo. Nothing here uses your mic. bolipi itself runs on your Mac.
Press play, or hold the key to drive it.
On your Mac you hold fn. Here, hold this key (or Space) to drive the demo.
How it works
Three steps. One key.
No app to switch to and nothing to click. bolipi works where you're already typing.
- 1
Hold fn
Press and hold the fn key in any app, wherever your text cursor is.
- 2
Speak and watch the words
Your words appear live in a small bubble by the cursor. Blue words are locked in; grey ones may still change.
- 3
Release
bolipi adds punctuation, drops the ums, applies your corrections, and pastes the text where your cursor was.
Features
Built for how people actually talk.
You think out loud, change your mind, and say um. bolipi handles all of it, right on your Mac.
Live words as you speak
Blue words are locked in. Grey words may still change as you keep talking.
Clean-up built in
Punctuation goes in, the ums and uhs come out, and “Friday, no wait, Thursday” becomes “Thursday”.
Works in every app
Slack, Chrome, Notes, Mail, VS Code, terminals: if you can paste there, bolipi can type there. Your clipboard is restored after.
Private by design
Speech recognition and clean-up run on your Mac. No account. No telemetry.
Fast
For typical dictations, clean text is ready about 1–2 seconds after you let go.
Runs on 8 GB Macs
Light enough for an 8 GB MacBook. It uses roughly 1.1–1.6 GB of memory while running.
Privacy
Your voice stays on your Mac.
Audio is never saved.
Your voice becomes text on your Mac. The audio never leaves the device and is never written to disk.
No account. No telemetry.
There's nothing to sign up for, and bolipi doesn't send anything about how you use it.
Never online.
The speech models come inside the app, so bolipi works with Wi-Fi off from the very first launch.
Install
Download. Drag. Then just talk.
Grab the app, drag it into Applications, and open it. bolipi is free during the beta.
What you need
- A Mac with Apple Silicon (M1 or newer)
- macOS 13 or later
- About 1 GB of disk. The speech models come inside the app
- English. It's the only language bolipi understands for now
Then, one-time setup
- Step 1
Drag bolipi into Applications
Open bolipi.dmg, drag the app into your Applications folder, then open it.
- Step 2
Allow Microphone
So bolipi can hear you while you hold fn.
- Step 3
Allow Accessibility
System Settings then Privacy & Security then Accessibility.
- Step 4
Set the Globe key to “Do Nothing”
System Settings then Keyboard then “Press 🌐 key to” then “Do Nothing”.
FAQ
Questions, answered.
Yes. Speech recognition and clean-up both run on your Mac. Your audio never leaves the device and is never saved to disk. The models come inside the app, so bolipi never goes online.
Any app where you can paste: Slack, Chrome, Notes, Mail, VS Code, terminals, and so on. bolipi pastes where your cursor is, then restores your clipboard.
Nothing. bolipi is free during the beta.
English only, for now.
Not yet. bolipi needs a Mac with Apple Silicon (M1 or newer) running macOS 13 or later.
No. You hold fn (Globe) by default, but you can pick another hotkey in bolipi's Setup, such as Right ⌥ or ⌃ Space.
Roughly 1.1–1.6 GB while it's running. It runs on 8 GB MacBooks.