These Smart Glasses Text Silent Mouth Movements

Subvocal’s smart glasses turn silent mouth movements into text, without requiring a voice. The Y Combinator-backed wearable reads subvocalised speech — lips and jaw shaping syllables silently — and sends the result to a screen. The company announced it on Tuesday, November 3, 2026.

What Exactly Do These Smart Glasses Text, and How?

The premise fits in a lift ride. Subvocalisation means forming words with your mouth and tongue without pushing air through your vocal cords — like the faint lip movements you make while reading a sentence in your head. The headset watches those movements, then turns them into typed input for a computer.

Subvocal has working prototypes, not just renders. In one demo video, Y Combinator partner Dave Lieb wears the hardware, tries it, and is impressed enough to offer funding on the spot. The unit in the clip is visibly 3D printed, and the company admits it isn’t what buyers will eventually get. Here’s the thing: a prototype that reads lips silently marks a research milestone, not a finished product. So who’s actually building it?

Who Is Subvocal, and Why Does Y Combinator Back It?

Subvocal is a startup in the Y Combinator cohort, pitching silent input rather than another camera-on-your-face gadget. That distinction matters. Most wearables in this category sell capture — photos, video, and notifications. Subvocal sells output: getting words from a person’s head into a machine, without a keyboard or microphone.

The company’s website shows mockups of a final design, far sleeker than the printed demo rig. Its pitch fits naturally alongside the wider wave of AI prescription glasses now crowding the eyewear market. No manufacturing partner, no ship window and no retail channel have been named. That gap between prototype and product is where most hardware startups die. Now, the numbers.

What Do We Actually Know About the Specs?

Almost nothing. That’s the honest answer as of November 2026. The company hasn’t disclosed processor specifications. It hasn’t shared RAM capacity, internal storage, or display specifications either. Battery life remains unknown, too.

Specification	Status as of November 2026
Processor	Not disclosed
RAM and storage	Not disclosed
Display	Not disclosed
Battery life	Not disclosed

Verdict: A funded prototype with zero published specs signals intent, not a buying decision.

Worth noting, the retail price hasn’t been announced.

What Happens Next for Silent-Speech Input?

The commercial case rests on places where talking is impractical: open-plan offices, hospital wards, factory floors, and libraries. Enterprise buyers already pay for this kind of workflow tooling. Vendors such as Cognizant, which is retraining thousands of staff around voice and AI interfaces, show how much money sits behind hands-free data entry. For more detail, see Gizmodo.

The risk is accuracy. Silent speech recognition has to work on the first attempt, in noisy rooms, for people with accents, dental work, and beards. A demo clip proves the concept; a shipping product must survive a Tuesday afternoon. Until Subvocal publishes silicon, “smart glasses text” remains a promise, not a product.

Beyond the Office: Where Silent-Speech Input Could Sell Next

Silent-speech input could also reach consumer markets, especially among people with speech impairments or those in noisy environments. Its versatility may appeal to sectors such as education and entertainment, expanding its potential market reach.

FAQs

What are the current specifications of the smart glasses?

As of November 2026, the smart glasses’ specifications remain undisclosed. That includes key details such as processor specifications, RAM capacity, internal storage, display specifications, and battery life. Without this information, potential buyers can only speculate about the product’s capabilities.

What is the potential market for silent-speech input technology?

Silent-speech input suits settings where speaking is impractical, including open-plan offices, hospital wards, factory floors, and libraries. Enterprise buyers already invest in these workflow tools, pointing to significant market potential. Companies like Cognizant are retraining staff to use voice and AI interfaces effectively. To succeed commercially, though, the technology must prove highly accurate in a range of challenging conditions.