Rokid First to Natively Add Gemini, ChatGPT to AR Glasses
AR glasses maker Rokid just released a major software update for its eyewear, becoming the first manufacturer to natively integrate both Google's Gemini and OpenAI's ChatGPT models. The update positions the glasses as a multi-modal AI assistant, pushing forward the integration of generative AI into consumer-facing embedded hardware.
This move positions Rokid's hardware not just as a content consumption device, but as an interactive platform for embodied AI. Founded in 2014, the company has focused on both enterprise and consumer applications, with previous industrial uses in sectors like oil & gas, aviation, and logistics. The company's flagship consumer products, like the Rokid Max 2, function as wearable displays, projecting a virtual 215-inch screen. They feature Micro-OLED displays with a 1920x1080 resolution per eye, a 120Hz refresh rate, and weigh only 75 grams. These glasses connect via USB-C to phones, laptops, and gaming consoles, essentially acting as a private, portable monitor. A key differentiator for Rokid is its "open ecosystem" approach, which contrasts with the walled-garden strategies of competitors like Meta and Apple. The company provides software development kits (SDKs) for its YodaOS (based on Android Open Source Project), fostering a developer community that is the largest of its kind in China, with over 20,000 developers. This software update leans into a "bring your own AI" model, allowing users to choose their preferred assistant instead of being locked into a single proprietary one. This strategy is also seen in their 38.5-gram, display-free "AI Glasses Style" which uses a dual-chip architecture to manage always-on tasks and more intensive AI workloads. The AR glasses market includes major players like XREAL, Microsoft with its enterprise-focused HoloLens 2, and Magic Leap. In the more consumer-focused smart glasses space, Rokid's primary competitor is Meta's Ray-Ban Stories, which are display-less and focus on camera and audio features. By integrating both Google's and OpenAI's models, Rokid avoids dependence on a single AI provider. This allows developers building on Rokid's platform to leverage the distinct capabilities of different foundation models for various robotics and automation-related applications, from real-time translation and object recognition to complex instruction overlays.
