Hands-free dictation: talk without holding a key
Push-to-talk dictation has a hidden requirement: a hand free to hold the key down. The moment your hands are full, or you are pacing around thinking out loud, that requirement is the whole problem.
Most dictation is push-to-talk. You hold a key, you talk, you let go, the text appears. It works, but it assumes you are sitting at the keyboard with a spare hand. Walk away from the desk, carry something, or just talk with your hands the way people actually think out loud, and push-to-talk stops being an option.
Hands-free mode is the other shape: start it once, and it keeps listening. Talk, stop to think, talk again - nothing to hold, nothing to re-press between sentences. It is one of the ways MightyMouse turns talking into typing in any Mac app - what you said gets typed in as you go, not all at once at the end.
How it starts and stops
There is no setting to turn on. Double-tap the listen hotkey and hands-free starts; a single tap of the same key stops it, the same way a tap already stops an ordinary dictation session. The gesture is the entire feature - you cannot leave hands-free half-armed the way you could a setting you forgot you switched on.
Once it is running, a card on screen says so, and every finished utterance gets its own small confirmation as it lands - proof the app heard that one and is still listening, not silence you have to trust.
The one problem hands-free has to solve
Push-to-talk never has to guess when you are done - you tell it, by letting go of the key. Hands-free has no key to let go of. Something else has to decide "that was the end of a sentence" versus "they are just thinking," using only the sound of a pause.
Too short a pause and it cuts you off mid-thought, splitting one sentence into two pastes. Too long and every natural breath between sentences reads as dead air, and the app sits there looking like it stopped working. There is no threshold that is right for every voice or every room - only one that has been tuned against real ones.
The listening itself runs on Silero VAD, a small model built for exactly this: telling speech from silence frame by frame, on your own machine, not a cloud call. What it decides gets fed to the same local Whisper transcription every other dictation mode here uses - the mic and the pause-detection are what is different about hands-free, not how the words get turned into text.
Each pause is its own paste
This is the part that separates hands-free from just "dictation without the key." Ordinary push-to-talk records the whole hold as one block and pastes it once, at the end. Hands-free pastes after every pause it detects as an ending - so a two-minute stream of thinking out loud arrives as several separate insertions, not one long paste you have to wait for.
That also means a pause you meant as an ending really does end the thought and get typed, before you have said the next one. You are not dictating into a buffer that reveals itself all at once when you finally stop.
Silence that is not actually speech - dead air, a cough, a mic pickup with nothing said - does not get pasted at all. Nothing heard means nothing typed, which matters more here than in push-to-talk: with no key marking where you meant to start and stop, the room between sentences is the only signal the app has, and it has to be right about which parts of it were not words.
What hands-free costs you
An honest list:
- It can still cut a sentence early. A pause tuned to feel natural for a fast talker will occasionally read as the end when you were only catching your breath. Push-to-talk cannot do this - it only ends when you let go.
- It is an always-on mic while it runs. Push-to-talk only listens while the key is held. Hands-free listens continuously until you tap to stop, which is the entire point but is worth knowing.
- It is a gesture, not a setting. There is nowhere to configure it on or default it open - you double-tap into it every time, which is deliberate but means it is never one click away.
What you get for that: no hand tied up holding a key, and a sentence gets typed the moment you finish saying it rather than waiting for you to remember to let go. For the broader case for dictating into any Mac app with transcription that stays on your machine, see the pillar this sits under.
Talk without holding a key down.
Free local transcription, into any Mac app. No card to try it.
Common questions
How does hands-free dictation know when I am done talking?
It listens for a pause and treats a long enough one as the end of what you were saying. There is no key to let go of, so the pause itself is the only signal - which is also why it can occasionally end a sentence a beat earlier than you meant.
Does hands-free mode send my voice to a server?
No. Both the pause detection and the transcription run on your own machine, the same as push-to-talk dictation. Nothing about going hands-free changes where the audio goes.
Can I turn on hands-free mode by default?
No, and that is intentional - it is a double-tap of the listen hotkey, every time, not a setting you leave on. A mic that can be armed and forgotten is worse than one you have to deliberately start.
If holding a key down is the part that bothers you: MightyMouse dictates into any Mac app with free local transcription, hands-free included - double-tap and start talking.