Talk to it, hands-free.

Built to be talked to, not typed at.

Tap the mic once and dictate. Your words appear in the input as you speak, in two tones so you can see what has been heard and what is still landing. Pause for a moment and the message sends itself — a small ring around the mic winds down so you always know what is about to happen. Say "send" to send sooner. Speak again to keep going.

When Presence is replying, the mic holds — not off, just resting — and picks back up the moment the response finishes. You can hold a whole conversation in the car, on a walk, mid-cooking, without ever touching the screen. Replies can be read back aloud, so it is a real two-way conversation rather than a screen you have to watch.

Hold the spacebar to talk, release to send — the same idea as a walkie-talkie. If you have already started typing, hold Shift and Space instead so you do not disturb what is in the box. Available in Chrome, Edge, and Safari; the mic stays quietly out of the way if your browser does not support it.

Here is the part most people get backwards: you do not need to compose a tidy prompt before you speak. Years of search boxes trained you to boil a thought down to a clean instruction first. Resist it. Presence works better on the messy, half-formed, thinking-out-loud version — the way you would explain something to someone who already knows the context.

Ramble, backtrack, say "actually, no, the other thing", leave the asides in. All of it is signal: it tells Presence what matters to you, where you are unsure, where the real question is hiding. A polished one-liner strips that away and leaves it guessing. Raw human speech, mess and all, is the richer input — the synthesis is its job, not yours.