At a time when AI technologies and smart wearables are converging at a faster pace, iFLYTEK has introduced another new development. Recent reports indicate that iFLYTEK AI Glasses will officially appear at the Canton Fair on April 15, becoming
one of the first smart-glasses products to integrate “AstronClaw.”
The so-called “AstronClaw” is an AI agent development and application tool built by iFLYTEK on March 12, 2026 based on the OpenClaw architecture. Its formal name is AstronClaw, and it has been described in the industry as iFLYTEK’s version of
“AstronClaw.”
With “one-click cloud deployment, sandboxed execution, and full access to mainstream IM platforms” as its core advantages, the product lowers the barrier to AI adoption while safeguarding data security, helping turn AI into a broadly accessible and efficient productivity tool.
By bringing this architecture into iFLYTEK AI Glasses, large-model capabilities are being extended from the software side to smart
hardware terminals for the first time.
As iFLYTEK’s latest commercialization in AI hardware, the AI glasses weigh only 40 grams, making them one of the world’s lightest multimodal smart glasses with binocular monochrome displays. Leveraging iFLYTEK’s core strengths in speech recognition, machine translation, and multimodal interaction, the glasses integrate full-scenario translation capabilities and introduce a multimodal noise-reduction scheme based on lip-motion recognition. By capturing the speaker’s lip movements with a camera and combining the signal with bone-conduction microphone input, the system can accurately lock onto the active speaker in complex multi-person environments and effectively filter ambient noise.
This reportedly boosts speech recognition and translation accuracy by more than 50 percent, significantly improving the practical experience in cross-language communication, meetings, and similar settings.
In addition, based on iFLYTEK’s self-developed AstronClaw architecture and ecosystem, the AI glasses focus on building a “Claw + hardware” super AI assistant.
This enables a complete chain spanning large-model invocation, ecosystem service access, and multimodal understanding and execution, moving GlassClaw from concept toward real-world deployment and further improving the integrated AI capability loop across software and hardware.
