OpenAI early investors are bullish! Brain-machine startup Sabi plans to ship by the end of this year with a “thinking-to-type” beanie

Sabi, a California neuroscience startup backed by Silicon Valley legendary venture capitalist and OpenAI early investor Vinod Khosla, has recently unveiled its wearable brain-computer interface product, claiming that all it takes is “thinking” to type on a screen, with the goal of launching its first consumer-style beanie hat by the end of 2026.

you can now control things with your brain. literally.

we’re building the most wearable BCI on the planet, with @sabicap, backed by @khoslaventures @accel @initialized & @kevinweil.

we collected the world’s largest neural dataset and trained the most capable Brain Foundation… pic.twitter.com/FIQiH7nIuL

— Rahul Chhabra (@rahulchhabra07) April 16, 2026

Think to type: Sabi turns imagination into reality

Sabi, a startup in Silicon Valley, recently announced that it is developing a wearable brain-computer interface device shaped like a beanie hat. It can decode the user’s inner language into text in real time, allowing users to simply “think” in their heads so the text appears directly on the screen. CEO Rahul Chhabra said the first product is expected to be launched on the market by the end of 2026, and a baseball cap version is being developed in parallel.

So-called brain-computer interface (Brain-Computer Interface, BCI) refers to technology that creates a direct communication channel between the brain and external devices. Major companies include Neuralink, Musk’s company, among others. The company has recently reportedly succeeded in enabling ALS (ALS) (amyotrophic lateral sclerosis) patients to speak again by implanting a neural-signal chip.

(Speak without speaking! Neuralink’s brain-chip breakthrough clears the way for speech-impairment treatment, bringing hope for ALS)

Sabi’s roadmap is more distinct: it not only focuses on avoiding invasive surgery, but its products will primarily target everyday consumers.

100,000 micro sensors break the bottleneck for wearable brain-computer interface devices

Sabi’s reading cap uses EEG (EEG) technology—recording brain electrical activity with metal sensor pads attached to the scalp. Because signals from wearable devices must pass through skin and the skull in order to be detected, their strength is far lower than that of implantable devices placed directly near neurons. Therefore, Sabi chooses to increase the density of its sensors:

Typical EEG devices are equipped with only dozens to hundreds of sensors, while Sabi’s hat will integrate as many as 70,000 to 100,000 micro sensors, enabling precise localization of the position and content of neural activity.

As for AI models, Sabi is building a “brain foundation model (Brain Foundation Model).” By training on large-scale neural data across users, it learns the patterns of brain activity associated with inner language. Currently, the company has accumulated brain data from 100 volunteers, totaling 100,000 hours.

Chhabra said that in the early stage, the device targets a typing speed of about 30 English letters per minute. Although this is lower than typical typing speeds, it will keep improving as users spend more time using it.

Key barrier for the wearable consumer market: ready to use, close to everyday life

JoJo Platt, an independent neuroscience consultant in San Francisco, said, “If wearable brain-computer interface devices are truly going to enter the consumer market, the product experience must meet higher standards.”

Most BCI devices must be recalibrated before each use, because users’ brainwaves change from day to day depending on fatigue and how focused they are. A consumer product must be designed to be plug-and-play and operate stably.

He also emphasized that appearance design matters just as much. Whether for medical or consumer use, users tend to choose devices that blend into everyday life without looking out of place—such as headphones.

In terms of business backing, Sabi is backed by Khosla Ventures founder and OpenAI early investor Vinod Khosla, who said, “The most important application of BCI is to let you communicate with a computer using your thoughts. If you want a billion people to use their thoughts to operate computers every day, then this technology can’t be invasive.”

Neural data is highly sensitive—privacy protection is paramount

The idea of thinking to type is undoubtedly impressive, but it also draws public attention to the privacy and security of neural data.

In response, Chhabra said that the process of uploading data from the device to the cloud uses end-to-end encryption throughout. Sabi’s AI models can be trained directly on encrypted data without needing to access raw brain data. In addition, the company also works with neural security experts from institutions such as Stanford University to conduct a comprehensive audit of the technical architecture.

Chhabra emphasized, “We understand that neural data is the most private data humans can have, and handling it with care is a basic respect for users.”

This article caught the eye of OpenAI’s early investors! The BCI startup Sabi ships the “think to write” beanie hat as early as the end of this year—the earliest mention appeared in Chain News ABMedia.