Dictation & Push-to-Talk
Push-to-Talk Dictation — Hold a Key, Speak, and It Types for You
Last Updated: 2026.07.08
Overview
Push-to-Talk lets you dictate straight into the chat box without ever leaving your keyboard. You hold a key, speak, and let go — Sapience transcribes what you said and drops the text right where your cursor is, ready to review and send.
It works just like a walkie-talkie: press and hold to talk, release when you're done. There's no button to hunt for and no separate window — your words appear in the message box you were already typing in.
This is the fastest way to get a long thought into a chat message. Speak a paragraph in the time it would take to type a sentence, glance at the text, and hit send.
How to Use It
- Click into the chat box so your cursor is there.
- Hold down the Alt key (the default) and start speaking.
- While you hold the key, a small red "Listening…" pill appears — that's your cue that Sapience is hearing you.
- Release the key when you finish your thought.
- The pill turns blue "Transcribing…" for a moment while your speech is turned into text.
- Your transcribed words appear in the chat box. Read them over, edit anything you like, and press send when you're ready.
That's the whole loop: hold → speak → release → text appears. You can do it again as many times as you like — each burst adds to whatever is already in the box.
Tip: Push-to-Talk inserts text; it doesn't send it. Nothing goes to the agent until you press send, so you always get a chance to read and tidy up first.
Choosing Your Key
The hold-to-talk key is Alt by default, but you can change it — or switch it off entirely — in Settings → Advanced.
OptionWhat it doesAltHold Alt to dictate (the default).CtrlHold Ctrl instead.Cmd / ⌘Hold the Command key (handy on a Mac).OffTurns Push-to-Talk off completely.
Why change it? A modifier key like Alt occasionally overlaps with a shortcut in your browser or operating system. If holding Alt does something unexpected on your machine, switch the dictation key to Ctrl or Cmd — or set it to Off if you'd rather not have a hold-to-talk key at all.
What Happens Behind the Scenes
You don't need to know the plumbing to use it, but here's what the "hold, speak, release" is actually doing:
- Pressing the key opens your microphone and starts recording — but only for as long as you hold it down.
- Releasing the key stops the recording immediately and sends that short audio clip off to be transcribed.
- The transcription comes back as text and is inserted at your cursor in the chat box.
- Nothing is saved. The audio is used only to make the text — it is not stored as a file in My Files, and the "listening" step doesn't clutter your conversation with extra messages. Push-to-Talk dictation is deliberately ephemeral: it exists just to get your words onto the page.
How It's Different From the Other Voice Tools
Sapience has more than one way to use your voice. Here's when to reach for each:
You want to…UseHowSpeak a message directly into the chat box, hands on the keyboardPush-to-Talk DictationHold Alt (or your chosen key), speak, releaseSpeak some text and paste it anywhere (a note, an email, another app)Dictate ModePress Ctrl+Shift+D for the floating panel → text is copied to your clipboardHave a real, back-and-forth spoken conversation where the agent talks backVoice ChatClick the Voice button
The big distinction: Push-to-Talk types into the chat box you're already in, while Dictate Mode hands the text to your clipboard for pasting elsewhere, and Voice Chat is a two-way spoken conversation rather than dictation at all.
Good to Know
- You're always in control of what's sent. Push-to-Talk only fills in the text — you review and send it yourself.
- It won't clog your files or your chat. No audio recording is saved to My Files, and no "I'm listening" messages appear in your conversation. Just the text you wanted.
- The pill tells you the state. Red "Listening…" means it's hearing you; blue "Transcribing…" means it's converting your speech to text. When the pill disappears, your words are in the box.
- It and the on-screen mic button take turns. Push-to-Talk and the blue microphone button in the composer are mutually exclusive — only one can be active at a time, so they never fight over your microphone.
- On an iPad or tablet, Sapience keeps your microphone permission "warm" between dictations so you aren't asked to grant mic access every single time. The trade-off is that your device's mic-in-use indicator may stay lit for a short while after you release the key — that's expected, and it stops on its own after a bit of inactivity.
Summary
Push-to-Talk Dictation is the quickest way to speak your message straight into the chat box: hold a key, talk, release, and your words are typed for you to review and send. It's Alt by default and fully adjustable — pick Alt, Ctrl, or Cmd, or turn it off, in Settings → Advanced. Because it types where your cursor is and saves nothing, it's the effortless middle ground between typing and a full voice conversation.