Subvocal AI Glasses Enable Silent Text Typing for $499

The Subvocal AI glasses have opened pre-orders at an early bird price of $499 ahead of an expected Q1 2027 rollout in the United States. Designed for professionals and commuters who need to take notes or message privately in quiet or crowded spaces, the device replaces audible speech with subvocal muscle-sensing technology. However, its advertised typing speeds remain unverified by third parties, and the final hardware specifications are still subject to change.

Translating Silent Speech into Text

Subvocal focuses specifically on environments where speaking aloud or typing manually is impractical. Instead of capturing acoustic voice data, the glasses detect the subtle neuromuscular activity generated when a user silently articulates words in their throat and face—a process known as subvocalization. An onboard AI model decodes these minute physiological movements into digital text in real time, allowing users to compose messages, draft notes, or interact with virtual assistants in complete silence.

RF Sensing Without Optical Cameras

From a hardware perspective, The Subvocal Company utilizes non-contact physiological signal sensing. Reports indicate that the mechanism leverages radio-frequency (RF) signals to monitor fine muscle movements. Notably, the hardware does not include optical cameras and does not rely on computer-vision lip-reading models, addressing key privacy concerns often associated with camera-equipped wearable devices. The company has not yet published its comprehensive sensor architecture.

Claimed Input Speeds and Personal Calibration

The startup claims its prototype hardware achieves silent text input speeds exceeding 200 English words per minute, roughly matching natural human conversational pace. These figures currently stem solely from internal laboratory benchmarks without external third-party evaluation. To maintain accuracy across different jaw shapes and facial structures, commercial versions will require users to complete a brief individual calibration setup before daily use.

Pricing, Backing, and Production Roadmap

Founded in 2026 by Georgia Tech alumni Aarush Yadav and Arnav Kumar, The Subvocal Company was accepted into the Y Combinator Fall 2026 accelerator program and has secured approximately $1.3 million in early funding from Y Combinator, Afore Capital, and other angel investors.

The product remains in active engineering development, and public renders only represent current conceptual industrial design. Standard retail pricing is planned at $699, while initial buyers can reserve a unit at the $499 promotional rate with a $199 refundable deposit prior to shipment.

Would you adopt subvocal muscle-sensing frames for silent messaging in public, or does non-contact RF sensing face too steep of an accuracy hurdle compared to standard keyboards?