While the smart glasses narrative has been dominated by American and Korean tech giants this year, a Chinese AI company just reminded the world that the wearable computing race isn’t a two-horse show. iFlytek, a company best known for its speech recognition and natural language processing technology, unveiled its latest AI-powered smart glasses at the BEYOND Expo 2026 in Macau — and they pack some genuinely clever features that set them apart from the usual spec sheet race.

Lightweight Build, Heavy on AI

At roughly 40 grams, the new iFlytek glasses are impressively light — about 20 percent lighter than comparable products in this category. That matters because the biggest barrier to smart glasses adoption isn’t FOV or battery; it’s whether people will actually keep them on their face. The aerospace-grade magnesium-aluminum alloy frame keeps things sturdy while the resin waveguide display and custom micro-optical module compress the optics down to something that doesn’t scream “prototype.”

But the real story here is what iFlytek’s speech-recognition heritage brings to the table.

Lip Motion Recognition Changes the Game

The standout feature is iFlytek’s self-developed lip-motion recognition multimodal noise reduction system. It’s a mouthful of a name, but the concept is elegant: a combination of 5+1 microphone array, cameras, and bone-conduction technology works together to identify who the wearer is looking at based on their lip movements. In environments that would make most speech-to-text engines throw up their hands — crowded exhibitions, train stations, airports — the glasses can isolate a single speaker and transcribe their words with surprising accuracy.

If you’ve ever tried to use real-time translation in a noisy coffee shop, you know how badly most solutions fail when ambient noise kicks in. iFlytek’s approach tackles the fundamental problem head-on by adding a visual signal (lip tracking) to the audio processing chain.

GlassClaw: The AI Agent on Your Face

The onboard AI agent, called GlassClaw, handles meeting transcription, information organization, email drafting, and complex multi-step workflows. During the demo, iFlytek’s General Manager Lin Huijie showed how voice commands could generate partnership proposals, organize travel itineraries, and send emails — all without touching a phone or laptop.

There’s also an intelligent teleprompter mode that auto-scrolls based on speaking pace, which could be genuinely useful for presentations, interviews, and public speaking.

The glasses support real-time translation across 122 languages, accents, and dialects, covering face-to-face conversations, phone calls, online meetings, and even on-screen text like menus and road signs through the AR visual translation mode.

The Price and Availability Question

Priced at 4,299 yuan (roughly $635), the iFlytek AI glasses aren’t cheap — but they’re also not operating at the premium tier of something like the Ray-Ban Meta or Samsung’s forthcoming Galaxy Glasses. Presales begin June 15 in China, and it’s unclear if or when international availability will follow.

The real significance here may be less about the device itself and more about what it signals. iFlytek’s ecosystem forum brought together Sunny Optical, Wanxin Optical, and Conant Optics to discuss industry standards for weight, comfort, and intelligence — suggesting a coordinated push to establish smart glasses as a legitimate product category in China rather than a niche experiment.

Takeaway

iFlytek’s AI glasses won’t be the device that makes or breaks the smart glasses category. But the lip-motion noise reduction system is the kind of clever, domain-specific innovation that bigger players haven’t thought to prioritize. As AI wearables multiply, the winners won’t necessarily be the ones with the biggest specs — they’ll be the ones that solve the actual friction points of wearing a computer on your face. Sometimes that means looking at how people speak, not just what they say.

Source: TechNode