Baidu and Alibaba may enter the “AI Glasses War”

Recently, dynamics have been frequent in the AR smart-glasses field. On October 16, INMO Technology took the lead in launching the new-generation AI smart glasses INMO GO3 and announced it would jointly build a native “AI+AR” content ecosystem with companies such as Tencent, Ant Group, and Zhipu AI. Close on its heels, Thunderbird Glasses will also release on October 23 the world’s first smart cinema glasses product that supports HDR technology.

I. Development history of AI glasses
AI smart glasses are the best carrier for landing AI applications. As wearable devices closest to the mouth, ears, and eyes—the three core human senses—they precisely match the core functions of the three senses: the mouth is responsible for language output, the ears undertake speech reception, and the eyes lead information intake. It should be known that 80% of human information comes from vision, which naturally makes glasses the optimal hardware carrier for AI and allows natural full-dimension input and output of sound, language, and vision. The development course of the smart-glasses industry is like a growth journey from tentative path-finding to full-speed sprinting, with four key stages that are clear and full of vitality.

2000–2013 was the period of technological accumulation, the industry’s “enlightenment moment”: in 2003, Vuzix launched the first smart device with basic display functions, bravely taking the first step in commercial exploration; in 2012, Google Glass brought a pleasantly surprising breakthrough by putting voice control and internet connectivity into glasses, directly building the “skeleton” for the functional design of subsequent products.

2014–2016 was the period of market-application expansion, with the industry entering the “acceleration lane”: a wave of capital poured in, intensifying competition among technical routes; the B-end market successfully achieved scaled rollout, with enterprise-grade products deeply cultivating industrial and medical scenarios, while consumer-grade products targeted fashion and entertainment needs—two tracks advancing in parallel.

2017–2022 was the period of precipitation and adjustment, the industry’s “cultivation period”: technical bottlenecks and high costs blocked the road ahead, the industry calmed down from the earlier feverish expansion and turned toward a “return to value.” Companies plunged into tackling core technologies and custom development; players such as Xiaomi and Huawei also came up with clever moves—using lightweight designs and people-friendly prices to push products into a broader lower-tier market.

Starting in 2023, the industry fully ushered in the “outbreak moment”: technologies such as optical display, 5G, and AI large models exerted force together, and AI glasses surged forward. Ray-Ban Meta, jointly launched by Meta and Ray-Ban, became a “phenomenon-level blockbuster,” galloping ahead in both sales and market share, and it even plans to expand capacity to the ten-million-unit level in 2026, setting off a lively “hundred-glasses war” across the entire industry.

II. Product categories of AI glasses
Currently, AI smart glasses fall into three categories: models without cameras, models with cameras, and models with displays, each with functional and technical emphases adapted to different scenarios.

Models without cameras focus on audio and wireless communication, featuring lightweight convenience and supporting basic services such as AI voice dialogue, music, and calls; models with cameras add shooting capability on top of the former, and combined with AI algorithms can achieve image recognition to expand scenario adaptation; models with displays are “AI+AR” fusion products whose core is AR optical-display technology, capable of projecting images in real time, and when paired with cameras can realize gesture interaction, thus offering better immersive experience and intelligent interaction. The main AI glasses currently on the market include:

III. The AI-glasses industry chain
The value of the AI-glasses industry chain is distributed in a smile curve, with added value concentrated in upstream core components and downstream brands/channels; mid-stream manufacturing has relatively low profit. Overall, it is similar to consumer electronics, and aside from brands, domestic industry-chain links such as hardware and software are relatively competitive.

Upstream core components and modules are key to performance and cost; apart from AR optical display, there are no obvious technical barriers. In terms of value, core computing chips (accounting for 20%–40%) and structural parts are dominant; AR glasses also contain an optical display module (accounting for 43% of cost).

The high-end chip market is dominated by Qualcomm, while domestic manufacturers are accelerating substitution in the mid- to low-end; optical-display modules are constrained by the low yield rate of waveguides and the difficulty of mass production of Micro-LED, with U.S. companies leading in technology.

The mid-stream consists of whole-device manufacturing by ODM/OEM firms such as Goertek, with layouts across the global supply chain.

The downstream is brand operations and channel sales: Meta leads the global market, while domestic players such as Xiaomi and Rokid are expected to open up the domestic market.

Alibaba’s Quark AI glasses are planned for release within the year and are currently in the mass-production ramp-up stage, with the sales date yet to be determined. Previously, Tmall leaked a CNY 9.9 (approximately USD 1.40) reservation page. The hardware of this glasses model adopts a dual-chip design of Qualcomm’s AR1 flagship chip plus a low-power coprocessor, paired with dual systems—Android and RTOS—to realize dynamic resource scheduling. Functionally, in addition to basic services such as voice assistant, music, and calls, it also integrates Alibaba’s ecosystem capabilities—relying on the Tongyi Qianwen large model and Quark AI, it supports Amap navigation, Alipay payments, Taobao price comparison, Fliggy business-travel reminders, etc., with ecosystem integration as its core selling point.

In addition, Baidu’s Xiaodu AI glasses are planned to go on sale within the year and previously appeared at the Baidu World Conference, featuring functions such as shooting and translation; ByteDance is exploring AI-glasses R&D, with the product still in an early stage and no release plan yet.

IV. Industry challenges for AI glasses
As an emerging product, AI glasses face significant challenges in the supply chain and production capacity, with multiple manufacturers encountering delivery difficulties.

The core bottlenecks are concentrated in three aspects: first, manufacturers’ deliveries are generally delayed. For example, the glasses Rokid released at the end of last year only saw the first batch delivered in June this year, and INMO Technology has also apologized for delayed shipments. Even though Rokid received more than 300,000 global pre-orders, it is still affected by capacity bottlenecks and faces considerable delivery pressure.

Second, there are two core problems in the supply chain: one is insufficient production capacity of core components and the inability to mass-produce modules at scale, leading to product supply shortages; the other is high production-process requirements—precise assembly in lightweight structures struggles to ensure yield, pushing up production costs.

Finally, against the backdrop of rapid market expansion, the supply chain must simultaneously cope with the dual pressure of technological iteration and scale ramp-up, placing extremely high demands on its allocation and rapid-adaptation capabilities.

[Disclaimer]: The above content reflects analysis of publicly available information, expert insights, and BCC research. It does not constitute investment advice. BCC is not responsible for any losses resulting from reliance on the views expressed herein. Investors should exercise caution.