your attempted speech
Setu Glasses
Nobody hears a thing.
Mouth the words on a crowded train or in a quiet office. Your message is sent, and your words stay yours.
Custom hardware and in-house models decode attempted speech. No voice required.
Introducing Setu Glasses
Say it without
a sound.
Message, write and talk to your AI just by mouthing the words.
Early access 2027 · Free to join
Concept render. Simulated message.
Setu makes wearables that understand what you mean. Say it without a sound. Control it without a touch. No typing. No tapping. Just mean it.
Choose how Setu
fits you
your attempted speech
neural interface
Setu Cap · The next step
Your hands stay free.
Choose, click and send while your hands do something else. Just mean it.
A non-invasive neural interface, built on our own hardware and decoding models. Nothing implanted. In development.
Setu
turn it on
Privacy
Private by design.
It’s off until you turn it on, a light shows when it is, and what you say is never used for ads.
On-device by design. Your data is never sold.
Designed to disappear.
Looks like glasses. Feels like glasses. Nobody needs to know.
Concept renders. Final design in development.
Every voice interface. Now silent.
Voice is the fastest way to write, to work with AI and to get things done. It’s also loud. Setu keeps all of its speed and none of the sound, so you can use it anywhere: an open office, a train, a library.
Email · Dictating
Hi Priya, the deck looks great. Two small edits before Friday: tighten slide 4 and add the pricing
Written silently
Dictate anywhere
Write emails, docs and messages at the speed of speech, even in an open office.
Your AI · Voice mode
Plan my week around the launch.Done. I moved two meetings and blocked Thursday for focus.Asked silently
Talk to your AI
All the speed of voice mode, without saying a word out loud.
Commands
Turn on do not disturbDoneRemind me at 6 to call MomSaid silently
Command anything
Every voice assistant and voice command, without the voice.
Where Setu is going
Less and less,
until it’s just you.
Every interface so far has asked you to do more: type, tap, speak. Setu is building the other way, one step at a time.
Say it without a sound.
Mouth the words. Nobody hears a thing.
Under the hood: custom hardware and in-house models that decode attempted speech. Our research →
Do it without a touch.
Choose, click and send just by meaning it.
Under the hood: a non-invasive neural interface that turns intent into action. Our research →
Think it. It’s done.
The words you mean, straight to the screen.
The goal: decoding intended language directly, with no movement, no voice and no surgery. Read our vision →
Every interface asked you to do more. Setu asks for less, until all it needs is you.
Be first on the journey.
Setu Glasses come first, with early access in 2027. Join the waitlist to hear first, and to follow everything that comes next.