Subvocal: glasses that let you type without making a sound
Reel 49 · explainer · November 3, 2026
Run a business and want this working for you, not on your to-do list?
I set Claude Code up inside one- to three-person businesses: your schedule, inbox and leads, done overnight and checked. From $149 a month, 3 founding seats at half price.
Subvocal is a YC startup (Fall 2026) that opened preorders on Oct 30, 2026 for glasses you type with by mouthing the words. Everything below is from Subvocal's own post, launch film and preorder page.
The explainer · 5 parts
- 01
Typing without a sound
Talking to your computer is fast, but you can't always speak out loud. Subvocal's idea: you mouth the words silently and the glasses type them.
In the company's words, it "decodes silent, almost invisible mouthing into text at more than 200 words/min."
Voice typing needs sound. Subvocal reads the mouthing instead, so the words arrive and the meter stays flat. The sentence is one from Subvocal's film. - 02
How it works: radar on your lips and jaw
The glasses send radar signals toward your face. When you silently form words, your lips and jaw move, and that changes how the signals bounce back.
The film labels three of them: upper lip, lower lip and jaw. The preorder page says Subvocal does not use or require cameras for silent speech input.
The glasses send radar signals toward your face. When your lips and jaw move, the way the signals bounce back changes. That change is the input. No camera is involved. - 03
AI turns the signals into text
Subvocal's AI learns from those returning signals and turns them into text. The film's example sentence is "Explain the new system to me."
- 200+ words a minute is Subvocal's own claim. No accuracy numbers have been published yet.
Subvocal's AI learns those signals and turns them into text. The sentence is the one in the film, shown at 200 words a minute. - 04
A conversation nobody hears
Besides messaging your team or telling an AI agent what to do in public, two people wearing Subvocal can have a private conversation, even across a crowded room.
Nobody in between hears a word.Two people wearing Subvocal can have a private conversation, even across a crowded room. Both lines are from Subvocal's film. - 05
What $499 gets you today
A preorder. The early-bird price is $499 (regular $699), with a $199 deposit per pair that's fully refundable any time before the final payment. Paid preorders are US only.
Shipping is expected to start in Q1 2027, delivery is subject to FCC approval, and no delivery date is confirmed yet.
$499$699Early bird · US onlyFrom Subvocal's preorder page. Taxes, shipping and prescription lenses cost extra, and no delivery date is confirmed yet.
Good to know
- Parts of the launch film are marked "*Prototype design", and the signal panel in it is a graphic. The prototype in the interviews is wired.
- Subvocal raised $1.3M from YC, Afore Capital and angels. That's the total, not YC alone.
- Not to be confused with AlterEgo (an MIT spin-out) or Cornell's EchoSpeech research.
- Sources: Subvocal on X, Oct 30, 2026 · the preorder page · Y Combinator · the launch film's subtitles
- An explainer by @copilot_shogo. Not affiliated with Subvocal.
Free for people building with Claude Code
Two of us run a six‑figure business on this Claude Code setup.
Most setups forget who you are every session. This one knows your goal, pushes back when you drift, and checks its own work. Take the starter version. It takes 15 minutes to set up.
✻
CLAUDE.md
- START-HERE.md
- CLAUDE.md
- prompts/interview.md
- agents/verifier.md
- diagnostic.md
The starter opens right away, and the link lands in your inbox to keep. One email a week after that, unsubscribe any time.
MORE LIKE THIS