Local voice-to-text
Local voice-to-text for the places you write.
hush·hush turns speech into organized text without making dictation a cloud workflow. It is made for the moment you are already writing.
Our recommendation
Use local voice-to-text when the input itself should stay close.
The strongest reason to use local voice-to-text is control. Your words are processed on your device, then inserted where you are already working.
- Local speech model
- Local organizing model
- Text lands where your cursor sits
One workflow, two local steps
The speech model turns what you said into text. The organizing model polishes structure, punctuation, and phrasing so the result is closer to what you meant to write.
Those model names stay generic because the user experience matters more than the vendor behind a model. The important part is that the work happens on your device.
Designed for repeated use
Local voice-to-text should feel quiet and repeatable. Hold the hotkey, dictate, release, and keep moving.
Smart Fix adds a memory for the words you correct. The product gets better at your names and terms without turning dictation into account sync.
Decision guide
Questions
What is local voice-to-text?
Local voice-to-text converts speech on your device instead of sending dictation to a remote service.
Why does hush·hush have an organizing model?
Natural speech is uneven. The organizing model turns spoken fragments, corrections, and long thoughts into text that is easier to use.
Does local voice-to-text work after setup?
Yes. After setup, daily dictation runs on your device. Licensing and update checks are separate from the act of dictating.