In 2025, China's smart glasses market has entered an era of "hundreds of glasses battling." Alibaba’s Quark launched its S1 and G1 series of AI-powered smart glasses, integrated with the Qwen large language model. Google announced Project Aura, a wired XR glasses collaboration with XREAL, set to launch in 2026. According to Counterpoint’s “Global AR Smart Glasses Brand Shipment Share Report for Q3 2025,” Chinese brand RayNeo by Rokid Innovation captured a 24% market share, maintaining its position as the global leader in the AR smart glasses market for two consecutive quarters.

The starting gun for smart glasses has fired: running in the right direction matters more than getting off to an early start.

At the 2025 Metaverse Innovation Conference, XREAL founder and CEO Xu Chi stated that smart glasses might be humanity’s “last screen.” He emphasized that seizing the opportunity presented by this “last screen” between the iPhone era and the advent of brain-computer interfaces requires building solid foundations and choosing the right direction—far more critical than simply rushing ahead.

However, for the smart glasses industry to truly advance to its next stage, it still depends on collaborative innovation across the supply chain. Liang Ziyu, head of Rokid’s Digital Culture Division, told The Paper that the entry of more major players into the sector is undoubtedly beneficial. The investment from tech giants will provide indispensable momentum. Only when smart glasses offer enough practical and engaging use cases will user engagement and sales increase, driving continuous product iteration and optimization.

At the 2025 Metaverse Innovation Conference exhibition area, attendees experienced smart glasses firsthand.

Powered by AI, smart glasses go mainstream

Earlier this year, Rokid founder Zhu Mingming demonstrated his “hands-free presentation assistant” by saying, “My speech script is right inside my glasses—I flip pages using a ring on my finger.” This vivid showcase helped smart glasses break out of niche circles.

Liang Ziyu told The Paper that the biggest change in the smart glasses industry this year has been its “mainstream breakthrough.” After Chinese New Year, people outside the industry began proactively asking about and wanting to try smart glasses. “People have realized there are one or two specific scenarios—like traveling abroad or giving speeches—where these glasses offer lasting utility.”

Reflecting on the past decade—from the VR wave to the metaverse hype and now the AI surge—Xu Chi noted significant progress in areas like device lightweighting and content richness. He believes the greatest opportunity for smart glasses in recent years lies in the practical implementation of AI, which has opened vast new possibilities for the industry.

Today’s AI systems are well-read but lack “eyes” to observe the world and “hands” to interact with it—they can only converse via text. Xu Chi argues that smart glasses could become an always-on interactive platform, enabling AI to act more proactively and personally. From a data perspective, large models are nearing saturation in their mining of existing datasets. Embodied intelligence and real-time data collected by smart glasses represent the essential path forward for AI’s next evolutionary leap—even toward achieving Artificial General Intelligence (AGI).

Xu Chi predicts that lightweight devices will increasingly dominate the market, with more smart glasses being developed in China. He envisions smart glasses as potentially “humanity’s last screen,” positioned between smartphones and future brain-computer interfaces. “We must seize this unique window of opportunity.”

He further suggests that in ten years, smart glasses may diverge into two distinct categories. One type will be relatively heavier and portable, aiming to replace monitors or tablets. These won’t be worn all day but will integrate PC and smartphone ecosystems to deliver smoother, smarter AI interactions. The other category will be ultra-lightweight, designed for all-day wear with the goal of replacing smartphones. Its core will be a powerful built-in AI assistant capable of continuously perceiving the environment—“seeing” what humans see and “hearing” what they say—to enable natural, real-time AI interaction.

Xu Chi demonstrating the first-generation spatial computing device.

Ningfeng Intelligence collaborates deeply with over ten domestic smart glasses companies. Its low-barrier, rapid-agent creation platform is specifically designed for metaverse content development. Ningfeng founder and CEO He Feixiang told The Paper that the rise of AI glasses is inevitable. Among future tech trends, the only standalone device capable of replacing smartphones as the next universal computing interface will undoubtedly be smart glasses.

“Current AI glasses primarily use green-tinted displays on lenses to deliver smart information—a transitional phase, in my view,” He Feixiang said. “In the next two to three years, AI and AR will deeply integrate, supporting full-color displays, spatial interaction, and spatial computing.” He predicts that as full-color display technology, spatial computing capabilities, processing power, and battery life continue improving, developers will create more engaging applications, enabling smart glasses to deliver richer visual experiences, boost usage frequency, and drive sales. Eventually, distinctions like “AI glasses” or “XR glasses” will fade away.

Challenges beneath the surface of prosperity

2025 has been dubbed the “Year One of the AI Glasses Boom.” According to research firm Counterpoint, global smart glasses shipments in the first half of 2025 more than doubled year-over-year, with Meta’s Ray-Ban smart glasses leading in volume. Despite the influx of players and heightened buzz, significant challenges remain hidden beneath the surface.

“Meta has lost tens of billions of dollars over the past few years, and there’s still no sign of a turnaround,” Xu Chi noted, drawing parallels between today’s smart glasses industry and the fragmented mobile phone market around 2005—characterized by diverse form factors and disjointed ecosystems. “Products are evolving rapidly this year, but user experience still lacks refinement. If the journey toward metaverse and AI terminals is a marathon, laying solid groundwork and heading in the right direction matter far more than sprinting ahead prematurely.”

He Feixiang added that for AI glasses to achieve all-day wearability and mass adoption, breakthroughs are needed in three key areas: battery life, computing power, and lightweight design. Current smart glasses typically last only 2–3 hours of continuous use—far short of the 7–8 hours offered by smartphones, which is necessary for true all-day wear. On the computing front, today’s devices have limited onboard processing power; future iterations must integrate stronger local compute capabilities to support more complex and imaginative functions.

Battery life, computing power, and lightweight design together form the “critical triangle” of AI glasses development. He Feixiang stressed that only when all three sides of this triangle are sufficiently extended can AI glasses mature and potentially replace smartphones.

The industry currently faces common technical hurdles. Liang Ziyu explained that whether it’s reducing weight, boosting computing performance, or extending battery life, solving these shared challenges requires supply chain collaboration. Within current technological constraints, manufacturers are actively building application ecosystems—adding features like payments, podcasts, ride-hailing, object recognition, and shopping—to enrich user experiences. Only when smart glasses offer abundant “usable and playable” scenarios will usage rates and sales rise, fueling product optimization. Simultaneously, companies must strike the optimal balance among technology, cost, and user experience to price products accessibly and unlock the mass market. The entry of more tech giants benefits the entire industry, as their investments bring crucial momentum. “Our focus is on how to collaborate with these giants to co-build ecosystems and achieve mutual success.”

Quark’s AI smart glasses offer navigation, translation, and other features.

When asked when the smart glasses industry might reach its “iPhone moment,” Xu Chi outlined four essential conditions:

First, hardware miniaturization—glasses, not bulky headsets, are the ideal form factor for all-day wear, and future smart glasses must weigh only tens of grams.

Second, robust native multimodal capabilities—AI must truly “see” the world.

Third, a unified ecosystem—the industry won’t be dominated by a single company but will require a global alliance. A fragmented ecosystem cannot support explosive growth; the sector needs an “iPhone-like” player to unify development and application standards.

Fourth, long-term memory—once smart glasses can continuously store and learn from years of a user’s meetings, experiences, and habits in a highly compressed, abstracted format, the onboard AI could evolve into a personal assistant that understands the user even better than they understand themselves.

“We’re boldly optimistic,” Xu Chi said. “In the next two or three years, we expect to witness a transformative shift in the smart glasses industry. 2027 or 2028 could mark the dawn of the all-day AI smart glasses era. We especially hope it’s 2027—that would be exactly 20 years after the iPhone’s 2007 debut, symbolizing the arrival of a new generation of computing terminals.”

In 2025, China's smart glasses market has entered an era of "hundreds of glasses battling." Alibaba’s Quark launched its S1 and G1 series of AI-powered smart glasses, integrated with the Qwen large language model. Google announced Project Aura, a wired XR glasses collaboration with XREAL, set to launch in 2026. According to Counterpoint’s “Global AR Smart Glasses Brand Shipment Share Report for Q3 2025,” Chinese brand RayNeo by Rokid Innovation captured a 24% market share, maintaining its position as the global leader in the AR smart glasses market for two consecutive quarters.

The starting gun for smart glasses has fired: running in the right direction matters more than getting off to an early start.

At the 2025 Metaverse Innovation Conference, XREAL founder and CEO Xu Chi stated that smart glasses might be humanity’s “last screen.” He emphasized that seizing the opportunity presented by this “last screen” between the iPhone era and the advent of brain-computer interfaces requires building solid foundations and choosing the right direction—far more critical than simply rushing ahead.

However, for the smart glasses industry to truly advance to its next stage, it still depends on collaborative innovation across the supply chain. Liang Ziyu, head of Rokid’s Digital Culture Division, told The Paper that the entry of more major players into the sector is undoubtedly beneficial. The investment from tech giants will provide indispensable momentum. Only when smart glasses offer enough practical and engaging use cases will user engagement and sales increase, driving continuous product iteration and optimization.

At the 2025 Metaverse Innovation Conference exhibition area, attendees experienced smart glasses firsthand.

Powered by AI, smart glasses go mainstream

Earlier this year, Rokid founder Zhu Mingming demonstrated his “hands-free presentation assistant” by saying, “My speech script is right inside my glasses—I flip pages using a ring on my finger.” This vivid showcase helped smart glasses break out of niche circles.

Liang Ziyu told The Paper that the biggest change in the smart glasses industry this year has been its “mainstream breakthrough.” After Chinese New Year, people outside the industry began proactively asking about and wanting to try smart glasses. “People have realized there are one or two specific scenarios—like traveling abroad or giving speeches—where these glasses offer lasting utility.”

Reflecting on the past decade—from the VR wave to the metaverse hype and now the AI surge—Xu Chi noted significant progress in areas like device lightweighting and content richness. He believes the greatest opportunity for smart glasses in recent years lies in the practical implementation of AI, which has opened vast new possibilities for the industry.

Today’s AI systems are well-read but lack “eyes” to observe the world and “hands” to interact with it—they can only converse via text. Xu Chi argues that smart glasses could become an always-on interactive platform, enabling AI to act more proactively and personally. From a data perspective, large models are nearing saturation in their mining of existing datasets. Embodied intelligence and real-time data collected by smart glasses represent the essential path forward for AI’s next evolutionary leap—even toward achieving Artificial General Intelligence (AGI).

Xu Chi predicts that lightweight devices will increasingly dominate the market, with more smart glasses being developed in China. He envisions smart glasses as potentially “humanity’s last screen,” positioned between smartphones and future brain-computer interfaces. “We must seize this unique window of opportunity.”

He further suggests that in ten years, smart glasses may diverge into two distinct categories. One type will be relatively heavier and portable, aiming to replace monitors or tablets. These won’t be worn all day but will integrate PC and smartphone ecosystems to deliver smoother, smarter AI interactions. The other category will be ultra-lightweight, designed for all-day wear with the goal of replacing smartphones. Its core will be a powerful built-in AI assistant capable of continuously perceiving the environment—“seeing” what humans see and “hearing” what they say—to enable natural, real-time AI interaction.

Xu Chi demonstrating the first-generation spatial computing device.

Ningfeng Intelligence collaborates deeply with over ten domestic smart glasses companies. Its low-barrier, rapid-agent creation platform is specifically designed for metaverse content development. Ningfeng founder and CEO He Feixiang told The Paper that the rise of AI glasses is inevitable. Among future tech trends, the only standalone device capable of replacing smartphones as the next universal computing interface will undoubtedly be smart glasses.

“Current AI glasses primarily use green-tinted displays on lenses to deliver smart information—a transitional phase, in my view,” He Feixiang said. “In the next two to three years, AI and AR will deeply integrate, supporting full-color displays, spatial interaction, and spatial computing.” He predicts that as full-color display technology, spatial computing capabilities, processing power, and battery life continue improving, developers will create more engaging applications, enabling smart glasses to deliver richer visual experiences, boost usage frequency, and drive sales. Eventually, distinctions like “AI glasses” or “XR glasses” will fade away.

Challenges beneath the surface of prosperity

2025 has been dubbed the “Year One of the AI Glasses Boom.” According to research firm Counterpoint, global smart glasses shipments in the first half of 2025 more than doubled year-over-year, with Meta’s Ray-Ban smart glasses leading in volume. Despite the influx of players and heightened buzz, significant challenges remain hidden beneath the surface.

“Meta has lost tens of billions of dollars over the past few years, and there’s still no sign of a turnaround,” Xu Chi noted, drawing parallels between today’s smart glasses industry and the fragmented mobile phone market around 2005—characterized by diverse form factors and disjointed ecosystems. “Products are evolving rapidly this year, but user experience still lacks refinement. If the journey toward metaverse and AI terminals is a marathon, laying solid groundwork and heading in the right direction matter far more than sprinting ahead prematurely.”

He Feixiang added that for AI glasses to achieve all-day wearability and mass adoption, breakthroughs are needed in three key areas: battery life, computing power, and lightweight design. Current smart glasses typically last only 2–3 hours of continuous use—far short of the 7–8 hours offered by smartphones, which is necessary for true all-day wear. On the computing front, today’s devices have limited onboard processing power; future iterations must integrate stronger local compute capabilities to support more complex and imaginative functions.

Battery life, computing power, and lightweight design together form the “critical triangle” of AI glasses development. He Feixiang stressed that only when all three sides of this triangle are sufficiently extended can AI glasses mature and potentially replace smartphones.

The industry currently faces common technical hurdles. Liang Ziyu explained that whether it’s reducing weight, boosting computing performance, or extending battery life, solving these shared challenges requires supply chain collaboration. Within current technological constraints, manufacturers are actively building application ecosystems—adding features like payments, podcasts, ride-hailing, object recognition, and shopping—to enrich user experiences. Only when smart glasses offer abundant “usable and playable” scenarios will usage rates and sales rise, fueling product optimization. Simultaneously, companies must strike the optimal balance among technology, cost, and user experience to price products accessibly and unlock the mass market. The entry of more tech giants benefits the entire industry, as their investments bring crucial momentum. “Our focus is on how to collaborate with these giants to co-build ecosystems and achieve mutual success.”

Quark’s AI smart glasses offer navigation, translation, and other features.

When asked when the smart glasses industry might reach its “iPhone moment,” Xu Chi outlined four essential conditions:

First, hardware miniaturization—glasses, not bulky headsets, are the ideal form factor for all-day wear, and future smart glasses must weigh only tens of grams.

Second, robust native multimodal capabilities—AI must truly “see” the world.

Third, a unified ecosystem—the industry won’t be dominated by a single company but will require a global alliance. A fragmented ecosystem cannot support explosive growth; the sector needs an “iPhone-like” player to unify development and application standards.

Fourth, long-term memory—once smart glasses can continuously store and learn from years of a user’s meetings, experiences, and habits in a highly compressed, abstracted format, the onboard AI could evolve into a personal assistant that understands the user even better than they understand themselves.

“We’re boldly optimistic,” Xu Chi said. “In the next two or three years, we expect to witness a transformative shift in the smart glasses industry. 2027 or 2028 could mark the dawn of the all-day AI smart glasses era. We especially hope it’s 2027—that would be exactly 20 years after the iPhone’s 2007 debut, symbolizing the arrival of a new generation of computing terminals.”