Features
Hands-free mode
Wake Infina with your voice and dictate without a key.
Hands-free mode lets you use Infina without touching your keyboard. Click where you want to type, say “type” and your message, then say “send it”. Say “open” and an app's name to open it. Anything else, like a website, needs “Infina” first. It's available on Mac, and on Windows laptops with at least 4 processor cores and 8 GB of memory — the speech model runs on the machine itself, so below that Infina says so rather than listening badly, and voice typing there needs the optional Cloud add-on. Which commands run once it's awake differs by computer — see Windows and Mac side by side. This page explains exactly how your audio is handled at every step, so you always know what stays on your device and what doesn't.
The short version
While Infina is asleep, the on-device listener is on and listening for you. It runs entirely on your computer and only listens for your wake words. Nothing is recorded or sent anywhere. The moment you wake it, Infina captures just that one spoken request and turns it into text right on your computer, then goes right back to sleep.
What to say
- “type can we move the call to 3pm”
- “send it”
- “open Slack”
- “Infina, open youtube.com”
type, send it and open plus an app work on their own. For everything else, say Infina first. After that the corner logo turns green, and while it's green you can skip “Infina”. All commands.
While asleep
When you're not actively using Infina, it sits in a sleeping state. A small model runs entirely on your device and listens only for your wake words. In this state:
- The listener is not recording and not sending anything anywhere.
- Audio is processed locally and immediately discarded — only the wake words can wake it up.
- Until you say a wake word, no audio ever leaves your computer.
When you wake it
As soon as a wake word fires, Infina captures just that one spoken request — the command or sentence you say next — and turns it into text right on your computer. For commands, the text is then turned into an action.
- Only your intended request is captured — never continuous or background audio.
- Infina detects when you've finished speaking and stops capturing — it doesn't keep an open microphone.
- Transcription runs on your computer. With the optional Infina Cloud add-on, your request travels over an encrypted connection to our cloud instead.
The wake and sleep cycle
Once your request is handled — or after a short period of silence — Infina returns to its sleeping, on-device-only state. The local listener takes back over, and nothing leaves your computer again until your next wake word.
A Stay awake setting controls how long Infina keeps listening after handling a request before it drifts back to sleep. If you don't say anything within that window, it returns to the on-device-only state automatically.
Your data and Privacy mode
By default, Privacy mode is on and nothing is kept: no dictation or voice-action text is saved. Transcription also runs on your computer by default, so your audio doesn't leave your device — with the optional Infina Cloud add-on, your request travels encrypted to our cloud providers instead. If you'd like a searchable history of what you've dictated, turn Privacy mode off in Settings → Data & privacy. Your stats, snippets, and dictionary are always kept either way.
You'll find Privacy mode in Settings → Data & privacy, and you can flip it on or off whenever you want. The trade-off: you won't have a searchable history of what you dictated.