Six ways in. One app underneath.
Bind a key, write a script, talk to it, float an overlay, put it on a schedule, or share it. All of it lives in one free app.
Press any combo, anywhere. It fires.
Keybind ships with zero shortcuts pre-assigned. You choose every combo deliberately.
Mic mute, window snapping, maximize, and AI Summarize come ready to bind. Nothing fires until you pick the combo.
KSL, a real language for when you want one.
Variables, control flow, functions, error handling: purpose-built for desktop automation, with a standard library across windows, clipboard, files, processes, HTTP, and AI.
What sets it apart from AutoHotkey and AutoIt: a capability-based permission model. A script can’t touch your files or network unless it declares it and you approve. Already have AHK scripts? Bring what you built.
// Prefix today’s screenshots with the date.
capability fs.write in "~/Pictures"
fn stamp(dir) {
let today = date.now("yyyy-MM-dd")
for f in fs.glob(dir, "Screenshot*.png") {
fs.rename(f, "${today}-${f.name}")
}
}Say it. It never leaves your machine.
Trigger any action by voice, fully on-device. Speech recognition runs locally. Your audio never touches a server.
Two modes: push-to-talk (hold a key, speak) or always-on. You define the phrases you want and map each to any action, so recognition stays fast and precise.
It learns how you talk: say a command your own way a few times and Keybind remembers it, on-device. Clear what it’s learned anytime.
HUDs that float over any app.
Build custom on-screen panels, pickers, and HUDs that float over whatever you’re doing. Drag components onto a preview of your screen, group them into named overlays, toggle each with a keybind.
Real native panels, not a web layer on top: sliders, toggles, and editable fields that stay sharp on any display.
Run anything, on a schedule.
An always-on scheduling engine: cron jobs, one-shots, run-on-startup, folder-watchers, or a custom condition — with retries, conditions, a dry-run mode, and in-app or webhook alerts.
With Pro, the Remote Scheduler dashboard runs and manages jobs on any of your machines from a browser.
Build it once. Share it.
Everything beyond raw key-remapping is a module: a small script that can add keybinds, overlay panels, or its own UI, installing live with no restart.
Build one, package it, publish to the marketplace, and buyers install with one click straight from the web into the app.
Snip a region, get the text.
Text Grab reads pixels, not apps. Select any rectangle on screen and the recognized text lands on your clipboard immediately, whatever’s underneath it.
That means it works on screenshots, video frames, PDFs, and locked-down apps that block copy. Anywhere your eyes can read text but a cursor can’t select it, Text Grab can. Fully offline, native Windows OCR. Free, in the core app.
Describe it. Get a script.
Type what you want in plain English. Keybind shows you a short plan, you approve it, and it writes a working script grounded in the real function library. Use your own AI key, or Keybind’s managed AI, no key needed.
on device.changed("headset") { audio.mute("system") process.launch("spotify") }
Summarize, too: highlight text anywhere, press a hotkey, get a clean TL;DR in a toast.