What is silent speech, and can you dictate without speaking aloud?

Silent speech is the recognition of words without any sound, using signals from the muscles that shape speech or from the brain. You cannot dictate that way on an ordinary computer or phone today. The sources we checked describe prototypes in research labs, and the speed measurements we found come from people with implants. The broader context is in our guide to the future of typing.

Silent speech reads the signals behind speech, not the sound

Speech is produced by muscles in the face, mouth and throat, and by the brain that drives them. A silent speech system skips the sound and reads one of those signals. There are two main approaches. One uses sensors on the face or throat that record muscle activity. The other uses electrodes implanted in the brain. Both aim to produce text or audible speech from words the person only mouths or intends to say.

Face sensors reached high accuracy in laboratory tests

The face-sensor studies report high accuracy without an implant. In the 2018 AlterEgo study, a wearable that reads silently mouthed words reached a median word accuracy of 92 percent in its user studies ¹. That is a median from the study's own user tests, under its own conditions, so it should not be read as an everyday accuracy rate.

In 2020, Gaddy and Klein trained a system on silent facial muscle signals and turned them into audible speech. In one data condition, the word error rate fell from 64 percent to 4 percent ². The gain depended on the condition, and the test was a laboratory dataset rather than everyday use. A reader should treat the 4 percent as the result of one condition, not as a typical error rate.

The speed figures come from implants, and none is a consumer device

The silent speech speed figure comes from a 2026 study of a brain-computer interface. In a copy task, the participant's silent speech was decoded at an average of 49.7 words per minute, with 96.5 percent accuracy ³. The same participant's vocalized speech reached 30.6 words per minute in that task ⁴, which makes the silent condition about 62 percent faster ⁵.

That is a real result for one person, who needed surgery to place electrodes in the brain. The 2026 study itself reports results from one participant with severe dysarthria, so the numbers do not describe the typical user ⁶. The brain-computer interface speed question covers the other implant studies in detail.

Can you dictate without speaking aloud?

Not with a product you can buy, as far as the sources we checked show. The face-sensor systems in the research need hardware on the face or throat, and the implant systems need surgery. None of the sources we checked describes a consumer product that dictates from silent speech, so a claim that a phone or computer can do this today should be checked against a published source before you rely on it.

For everyday writing, the practical answer is the keyboard. If you need to draft by voice but cannot speak aloud in the room, a typed workflow is the reliable choice today. A dictation app is useful when you can speak, and a keyboard is the input when you cannot. Our overview of how to dictate explains the setup for spoken drafting, and the speech recognition for speech impairments guide covers recognition for people whose speech is affected.

Why silent speech is hard to make accurate

Silent speech is harder to read than sound. A sensor on the face has to separate the words from the other movements of the muscles, and a brain electrode needs surgery and picks up signals from one person's brain. That is why the results above depend so much on the condition, the sensor and the participant. The studies also use different tasks and measures, so their figures are not directly comparable.

What would change the answer

Three signs would show that silent speech is moving toward everyday use: a study that reports results across more than one participant, a system that works without surgery or dedicated sensors, and a commercial product with published accuracy figures. Until then, a page that says you can dictate silently with an ordinary phone should be treated with caution.

For the other side of the comparison, see will voice replace keyboards and are smart speakers still popular. The term word error rate explains how the accuracy figures above are counted, and dysarthria explains the speech disorder in the 2026 study.