Talk.
Don’t click.
A Chrome extension that lets you operate any website with your voice. Click, fill forms, run multi-step tasks by saying them out loud, the way you’d describe them to a friend.
From a sentence to a screen full of stuff done.
Hold the shortcut. Talk like you'd describe a task to a coworker. No wake word. No commands to memorize.
Audio streams to gpt-realtime. It hears your intent, reads the page, and decides the exact actions to take, in milliseconds.
The extension clicks, types, scrolls, and submits on the real page, in front of you. Cancel anytime by saying stop.
Everything your mouse can do.
Plus thinking.
No new app. No fragile site integrations. The extension drives the same DOM you do. It’s listening, not waiting for clicks.
Buttons, links, dropdowns, dialogs, menus, infinite-scroll lists. If you can click it, you can say it.
Names, addresses, multi-step checkouts. Pause to read out the part you forgot. Resume mid-sentence.
“Search, sort by price, open the second one, add to cart, checkout.” One breath. Done.
No site integration. No API. The extension operates the DOM exactly like you would, anywhere.
You see every action happen on the page. Say stop or release the shortcut and it halts instantly.
Streaming audio in, intent out. The hand-off feels like talking to a colleague at the next desk.
The mouse turned sixty last year. Your hands are tired of it. Your eyes are tired of hunting for the button. Just say what you want, and watch the web do it.