Meta市值1.7万亿、XREAL累计亏损20亿,中国AI眼镜被低估了吗?Meta's Market Cap Hits $1.7 Trillion, XREAL Cumulative Losses Reach 2 Billion Yuan: Are China's AI Glasses Undervalued?


AI free token/AI免费积分

本文来源:https://mp.weixin.qq.com/s/LbLEaNbE_FFuBLT_8IYX7g
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。

 

图片

来源:朝阳资本论

作者:江夏

 

 

端侧AI革命以来,AI眼镜是PC产品之外最被寄予厚望的选择。

据《朝阳资本论》不完全统计,仅今年上半年,就有超过65款AI眼镜发布,其中雷鸟一家就发布了5款,包括雷鸟X3 Pro、雷鸟Air4 Pro、雷鸟GT Max、雷鸟GT、雷鸟V4等,售价从1849-8499元不等。

 

实际上,距离国内首款AI眼镜诞生,也不过才两年多时间。

彼时,早期玩家乐奇Rokid、雷鸟、XREAL,虽然早有深耕AR技术,但真正具备大规模量产“语音/视觉AI+显示/拍摄”的AI眼镜产品标准形态,是在2024年(XREAL的ONE系列)。

也就是说,国内AI眼镜发展,满打满算也就不到3年。

而海外首家Meta第一款智能眼镜为‌Ray-Ban Stories‌,于‌2021年9月‌正式发布;如果指具备实时AI助手交互能力的型号,则为‌2023年9月‌发布的‌Ray-Ban Meta‌(第二代)。

就是这样一个诞生不到3年的年轻赛道,即将出现AI眼镜“第一股”。

2026年4月1日,XREAL正式向港交所递交招股书。

紧随其后,爱企查信息显示,Rokid完成股份制改造,内部上市窗口指向7月;雷鸟创新则在年初拿下中国移动、中国联通超10亿元融资后,公开招聘具备上市经验的投资人总监。

 

三家头部企业几乎同时叩响资本市场的大门。

2026年,被行业称为“AI眼镜IPO元年”。

 

喧嚣背后,回到商业本质,AI眼镜赛道,真的是一门好生意吗?

 

图片
为什么是2026年?

 

IDC报告显示,2026年一季度,全球智能眼镜市场出货量达到356.6万台,同比增长130.1%。

 

具体到中国市场,一季度中国智能眼镜出货量61万台,同比增长23.5%,位居全球第三。

IDC预测,2026年中国智能眼镜市场出货量将突破491.5万台,扣除掉洛图科技预测上半年的101.2万台,今年还剩390.3万台缺口。

图片

为什么会在2026年爆发?

毕竟AI眼镜并非新生事物,谷歌早在2012年就发布了Google Glass。

说千道万,还是要归功于AI。

第一,AI大模型催熟。无论是语言还是视觉交互,2023年之后,大模型技术突破让AI眼镜从“能看见”变成“能理解””。科技公司关注的不是一副能显示信息的眼镜,而是新的AI入口。这一轮竞争背后,是模型能力、终端入口和生态控制权的重新分配。

 

第二,端侧AI革命。AI端侧设备的崛起并非简单的技术延伸,而是AI应用场景从云端向终端渗透的必然结果。“轻量化+AI”已成为明确趋势,智能眼镜重量小于50克的产品占比持续提升。IDC预测,2026年重量小于50克的智能眼镜产品占比将超过45%。

 

第三,政策红利释放。2026年,智能眼镜首次被纳入国家补贴范围,与手机、平板、智能手表并列成为四大补贴品类。按售价15%补贴、单件上限500元,单价不超过6000元的产品均可享受。补贴直接拉动了渠道备货和终端需求。

 

那么,业内具体几家究竟卖的如何

根据当前的市场格局来看,不同统计口径下排名各异。

从AR眼镜细分市场看,雷鸟创新以全球24.7%的市场份额、中国31.7%的份额位居第一。在更前沿的波导AR市场,Rokid以41%的份额位居全球第一。

从整体智能眼镜出货量看,Counterpoint数据显示,中国市场小米以28%的份额排名第一,阿里巴巴以21%位居第二。

从线上零售额看,奥维云网数据显示,2026年第一季度Rokid以1.2亿元、21.2%的份额居首。

从全渠道销量看,沙利文认证阿里巴巴旗下千问AI眼镜获得2026年1-5月“AI眼镜全国销量第一”。

简单来说,AR技术派(雷鸟、Rokid、XREAL)深耕技术和显示体验;跨界巨头派(小米、阿里)携生态和渠道优势迅速冲入前列。

 

但无论哪一派,目前都还没有谁能建立起绝对统治力。

图片
到处都是“第一”
赚钱了吗?

 

终端品牌打得不可开交,上游供应链“卖铲人”先赚得盆满钵满。

毕竟相对于雷鸟、乐奇们要出货,上游厂商是最先拿到实实在在的订单的。

 

一般而言,在一副带显示功能的AI眼镜中,光学显示部件成本占比达到四成至五成,主控芯片则占两成至三成,两者合计吃掉了产品七成以上的成本(东吴、方正券商研报数据)。

于是,市场先给出了答案。

Wind数据显示,AI眼镜概念板块今年以来持续受到资金追捧,万得AI眼镜主题指数(866535.WI)5月涨幅达17.71%。聚焦ePOP嵌入式存储的佰维存储、聚焦光学显示的蓝特光学涨幅都超过100%。

产业层面的订单同样在放量。

华为鸿蒙AI眼镜由歌尔股份代工;佰维存储为Meta、Google、小米等多品牌供应存储芯片;蓝思科技下半年批量交付光波导镜片;台湾光学厂今国光上半年营收同比增长58.78%,玉晶光上半年累计营收118.23亿元。

正如行业人士所言:“AI眼镜的硬件基本掌握在供应链,现在赚钱的主要是产业上游。”

 

这样的的历史规律,一次次在产业上重演。

智能手机爆发期,无论哪家手机品牌胜出,高通和康宁都稳赢;新能源汽车时代,宁德时代的市值超过多数整车厂。

终端的牌局越混乱,上游的筹码越稀缺。

和上游增长确定性相反,终端品牌是陷入了高增长、高亏损的囚徒困境。

图片

图源:XREAL招股书

 

从XREAL的招股书来看,营收在增长,但远未盈利。 

2023至2025年,XREAL营收从3.9亿元增长至5.16亿元,2025年同比增幅30.8%。毛利率从18.8%提升至35.2%,接近成熟硬件企业水平。

但同期净亏损分别为8.82亿元、7.09亿元和4.56亿元,三年累计亏损超过20亿元。即便剔除优先股公允价值变动等非现金项目,经调整净利润依然为-4.37亿元、-3.75亿元、-2.5亿元。

 

截至2025年末,XREAL现金及现金等价物仅约6363万元。而同期经营活动所用现金净额为-2.03亿元。流动负债净额从2023年的19.32亿元增至2025年的30.84亿元。

图片

图源:XREAL招股书

亏损的原因,其实摆在明面上。

一是优先股等公允价值变动带来的非现金亏损;二是持续的研发投入,三年分别约2.16亿元、2.04亿元、1.83亿元;三是作为新兴行业必须承担的营销和市场教育成本。

XREAL如此,乐奇Rokid也不例外,从成立至今同样经历了漫长的亏损期。

行业普遍共识是,智能眼镜是一个极其烧钱的赛道,超小光机、波导镜片、空间感知算法、高密度电池,每一项技术背后都需要巨额的“前期成本””。

XREAL之所以率先冲刺IPO,或许部分来自外部股东压力。

公开数据显示,XREAL至今累计融资超23亿元,投资方包括阿里、快手、红杉中国等明星机构。创始人徐驰合计持股约27.98%,虽为单一最大股东,但持股比例未过半数。

对于一家成立9年、累计亏损超20亿、现金流不足1亿的企业而言,IPO不仅是获取资金的渠道,也是投资人实现退出的优先路径。

乐奇Rokid和雷鸟面临类似的压力。

爱企查信息显示,2026年1月完成C+轮融资,新增股东包括蓝思科技、商汤科技、三星医疗等产业资本,现金流储备超过15亿元。

雷鸟则在年初完成中国移动、中国联通超10亿元融资。

“AI眼镜第一股”很可能在2026年内诞生,XREAL已递表,乐奇Rokid、雷鸟虽未正式递表但市场持续关注。

现在的问题不在于“是否上”,而是“谁先上”。

图片
Meta市值1.7万亿美元
中国AI眼镜是否被低估?

 

要向资本市场伸手要钱,总得拿出货真价实的筹码。

目前来看,AI眼镜赛道基本面还是扎实的。

首先,市场规模虽然不大,但增长潜力还是有,且目前并没有出现绝对头部,谁都有可能做第一。 IDC预测,2026年全球智能眼镜出货量将突破2368.7万台;艾媒咨询数据显示,2025年全球智能眼镜市场规模为125.8亿美元,预计2029年将达到1387.3亿美元。

 

其次,技术拐点已至。

 

今年,AI眼镜技术从单一功能向多形态分化,录音、拍摄、翻译、支付等功能落地刚需场景。

此外,巨头集体押注。

 

近两个月,乐奇Rokid、雷鸟、XREAL、阿里千问、科大讯飞等厂商密集发布或预告新品。Meta计划2026年底前将Ray-Ban Meta年产能翻倍至2000万副。

只是,“好赛道”不等于“好生意”。

 

当前AI眼镜面临的核心矛盾是,市场规模在爆发,但单个企业仍未找到盈利模式。

 

行业机构统计显示,行业新增的200多个AI功能,用户长期使用率不足6%。京东、天猫等主流电商平台AI眼镜平均退货率高达30%。

图片

图:Meta市值

国内的“AI眼镜第一股”,如果需要有参考标的,那么就是海外市场的Meta,毕竟Meta海外一家独大。 

截至目前,Meta美股市值1.7万亿美元。

这1.7万亿美元,对应的出货量和市场份额是:Meta在全球无显示智能眼镜市场的份额已接近84%。2025年其AI眼镜出货量达740万台,年增281.3%(Counterpoint数据),美国仍是全球最大智能眼镜市场。

至于中国市场,以乐奇Rokid祝铭明本人对Rokid的销量目标规划来看,2026年超过100万台,2027年达到200万至300万台,2028年突破1000万台。

也就是说,要用3年时间,来追平全球第一的出货量,而3年后,Meta又到了何方,尚未知。

值得一提的是,竞争格局方面,国内市场与海外市场高度集中形成鲜明对比,一年时间,市场前五的名单已发生剧烈变动。

这种格局意味着国内市场仍有变数,但也意味着没有企业能靠单一优势建立护城河。

谁能和Meta一拼?

基于当前的竞争态势,国内玩家大致可以分为四个梯队。

图片

第一梯队:雷鸟创新,属于技术派的“光学卷王”。

 

雷鸟是少数具备核心光学元器件全链路自研能力且能大规模量产的厂商,其优势在于光学技术的物理壁垒,上游供应商的护城河是物理极限,而物理极限是最顽固的。不过,雷鸟产品价格锚定中高端,难以触达大众市场且AI模型依赖外部供应。

图片

第二梯队:乐奇Rokid,跨生态“外交家”。

 

乐奇Rokid最鲜明的标签是开放,智能眼镜可自由切换千问、DeepSeek、Gemini等主流大模型。密集引入产业链企业作为股东,将供应商转化为利益共同体。先发优势积累的品牌声量和与央视世界杯的合作是差异化亮点,但盈利能力依然是市场关注焦点之一。

图片

第三梯队:阿里千问,生态派的最强选手。

 

千问AI眼镜深度打通支付宝、淘宝、高德等生态。这种“硬件+AI+生活服务”三位一体的模式,是互联网巨头做硬件的典型打法。2026年1-5月全渠道销量第一的认证证明了生态的力量,但挑战在于硬件并非阿里的核心能力,产品定义和供应链管理仍需时间打磨。

图片

第四梯队:小米、华为,蓄势待发的巨头。

 

小米以28%的出货量份额位居中国第一,华为首款鸿蒙AI眼镜起售价2499元。两者的共同优势是成熟的渠道网络、庞大的用户基数、深厚的供应链管理能力。一旦AI眼镜进入“性价比”和“生态协同”的竞争阶段,后劲不容小觑。

图片

至于科大讯飞,4299元的定价和122种语言翻译的定位,主要是切入商务翻译的垂直场景,在AI眼镜的泛大众市场竞争中,暂时还难以与前四家正面抗衡。

整体来看,行业1-2年内还是处于“百镜大战”的混战阶段,最后还是要看剩者为王。

乐奇Rokid创始人祝铭明在2026生态及开发者大会上也表示,“现阶段行业无需过度追逐短期市场占有率,整个赛道仍处于'黑莓时代',产业真正规模化爆发的临界点尚未到来。”

但是参考智能手机发展轨迹,2009年诺基亚、黑莓、HTC、摩托罗拉群雄并起,最终只剩苹果、三星、华为、小米等少数玩家,AI眼镜大概率会复刻这一剧本。

上游供应链的成熟将降低进入门槛,但品牌、生态和渠道的壁垒会越来越高。

最终能活下来的,可能不超过5家。

当AI眼镜真正成为“下一代计算平台”,厂商竞争的本质是模型能力、终端入口和生态控制权的重新分配。

拥有大模型能力、操作系统、硬件制造能力和渠道覆盖的企业,将更有机会走到最后。

 

注:文章图片来自网络,侵权请联系删除。
图片
图片

Momenta港股上市,700亿市值凭什么?

2026年的咖啡加盟商:不玩了、退钱!

三次港股递表,累计亏损超8亿,景泽生物IPO等得起吗?

图片

 

Source: https://mp.weixin.qq.com/s/LbLEaNbE_FFuBLT_8IYX7g If you have any copyright concerns, please contact: 13951431913#139.com or submit a comment on the right.

图片

Source: Chaoyang Capital Theory

Author: Jiang Xia

Since the edge AI revolution began, AI glasses have become the most promising choice besides PC products.

According to incomplete statistics by Chaoyang Capital Theory, in the first half of this year alone, more than 65 AI glasses were launched, with Thunderbird (Rajiao) releasing 5 models—including the Thunderbird X3 Pro, Air4 Pro, GT Max, GT, and V4—priced between 1,849 and 8,499 yuan.

In fact, it has been little more than two years since China's first AI glasses hit the market.

At that time, early players like Rokid, Thunderbird, and XREAL had long been deeply involved in AR technology, but truly scalable "voice/visual AI + display/capture" AI glasses only emerged in 2024 with XREAL's ONE series.

In other words, China's AI glasses industry has developed for less than three years at most.

Overseas, Meta's first smart glasses—Ray-Ban Stories—were officially launched in September 2021; the models with real-time AI assistant interaction capabilities came later, with Ray-Ban Meta (second generation) released in September 2023.

It is in this young track that has been around for less than three years that the first AI glasses "IPO" is about to emerge.

On April 1, 2026, XREAL officially filed its prospectus with the Hong Kong Stock Exchange.

Following shortly after, Ai企查 information showed that Rokid completed its shareholding reform, with an internal listing window pointing to July; meanwhile, Thunderbird Innovation secured over 1 billion yuan in financing from China Mobile and China Unicom earlier this year and publicly recruited an investment director with listing experience.

The three leading companies are almost simultaneously knocking on the door of the capital market.

2026 has been dubbed by the industry as the "year of AI glasses IPOs."

Behind the buzz, returning to the essence of business: is the AI glasses track really a good business?

图片

Why 2026?

According to an IDC report, global smart glasses shipments reached 3.566 million units in Q1 2026, a year-on-year increase of 130.1%.

Specifically for the Chinese market, Q1 smart glasses shipments reached 610,000 units, up 23.5% year-on-year, ranking third globally.

IDC predicts that China's smart glasses market shipments will exceed 4.915 million units in 2026. Deducting the 1.012 million units expected in the first half of the year from Lotu Technology's forecast, there remains a shortfall of 3.903 million units for the rest of the year.

图片

Why the explosion in 2026?

After all, AI glasses are not a new concept; Google launched Google Glass as early as 2012.

In the end, credit goes to AI.

First, AI large models have matured the sector. Whether through language or visual interaction, breakthroughs in large model technology after 2023 transformed AI glasses from "capable of seeing" to "capable of understanding." Tech companies are no longer focused on glasses that merely display information, but on new AI entry points. This round of competition is essentially a reallocation of model capabilities, terminal access, and ecosystem control.

Second, the edge AI revolution. The rise of edge AI devices is not a simple technological extension but an inevitable result of AI application scenarios penetrating from the cloud to the terminal. "Lightweight + AI" has become a clear trend, with the share of smart glasses weighing under 50 grams continuously increasing. IDC predicts that in 2026, smart glasses weighing under 50 grams will account for more than 45%.

Third, policy dividends are being released. In 2026, smart glasses were included in the national subsidy program for the first time, listed alongside smartphones, tablets, and smartwatches as one of four subsidized categories. With a 15% subsidy capped at 500 yuan per unit, all products priced under 6,000 yuan are eligible. The subsidy directly drove channel inventory buildup and终端 demand.

So, how exactly are specific companies in the industry performing?

According to the current market landscape, rankings vary depending on different statistical calibers.

From the perspective of the AR glasses细分市场, Thunderbird Innovation leads globally with a 24.7% market share and 31.7% in China. In the more cutting-edge waveguide AR market, Rokid holds a 41% share, ranking first globally.

From overall smart glasses shipments, Counterpoint data shows Xiaomi leads China with a 28% share, followed by Alibaba at 21%.

From online retail sales, Avic.cn data shows Rokid led in Q1 2026 with 120 million yuan in sales and a 21.2% share.

From omni-channel sales, a Frost & Sullivan certification confirmed that Alibaba's Qianwen AI glasses achieved "National Sales No. 1 in AI Glasses" from January to May 2026.

Simply put, the AR technology faction (Thunderbird, Rokid, XREAL) has deeply cultivated technology and display experience, while the跨界 giant faction (Xiaomi, Alibaba) has rapidly surged to the forefront leveraging ecosystem and channel advantages.

But regardless of which faction, no one has yet established absolute dominance.

图片

Everywhere there are "firsts," but are they making money?

[39] Terminal brands are locked in fierce competition, while upstream supply chain "shovel sellers" are reaping substantial profits first.

Terminal brands are engaged in fierce competition, while upstream supply chain "shovel sellers" reap hefty profits first.

After all, compared to companies like Thunderbird and Rokid needing to ship products, upstream manufacturers receive tangible orders first.

Generally speaking, in a pair of AI glasses with display functionality, optical display components account for 40% to 50% of the cost, while the main control chip accounts for 20% to 30%, together consuming over 70% of the total product cost (according to research reports from Dongwu and Founders Securities).

Thus, the market has given its answer first.

Wind data shows that the AI glasses concept sector has been continuously favored by capital flows this year, with the Wind AI Glasses Theme Index (866535.WI) gaining 17.71% in May. Companies focused on ePOP embedded storage like Biwei Storage and those focused on optical displays like Lante Optics both saw gains exceeding 100%.

Industry-level orders are also scaling up.

Huawei's HarmonyOS AI glasses are manufactured by Goertek; Biwei Storage supplies memory chips to Meta, Google, Xiaomi, and other brands; Lens Technology will mass-produce and deliver optical waveguide lenses in the second half of the year; Taiwan's optical manufacturer Jingguo Guang reported a 58.78% year-on-year revenue increase in the first half, while Yujing Guang accumulated revenue of 11.823 billion yuan in the first half.

As industry insiders put it: "The hardware of AI glasses is essentially controlled by the supply chain, and the main players making money right now are upstream of the industry."

This historical pattern has repeated itself time and again across industries.

During the smartphone boom, whether it was Apple or Samsung that won out, Qualcomm and Corning always profited; in the new energy vehicle era, CATL's market cap surpassed most automakers.

The more chaotic the terminal game, the scarcer the upstream chips.

In contrast to the certainty of upstream growth, terminal brands are trapped in a dilemma of high growth and high losses.

图片

Source: XREAL Prospectus

From XREAL's prospectus, revenue is growing, but profitability is far off.

From 2023 to 2025, XREAL's revenue grew from 390 million yuan to 516 million yuan, a year-on-year increase of 30.8% in 2025. Gross margin rose from 18.8% to 35.2%, approaching the level of mature hardware companies.

However, net losses during the same period were 882 million yuan, 709 million yuan, and 456 million yuan respectively, totaling over 2 billion yuan in cumulative losses across three years. Even excluding non-cash items such as fair value changes in preferred shares, adjusted net profit remained at -437 million yuan, -375 million yuan, and -250 million yuan.

As of the end of 2025, XREAL's cash and cash equivalents totaled only about 63.63 million yuan, while cash used in operating activities amounted to -203 million yuan during the same period. Net current liabilities increased from 1.932 billion yuan in 2023 to 3.084 billion yuan in 2025.

图片

Source: XREAL Prospectus

The reasons for losses are actually obvious.

First are non-cash losses from fair value changes in preferred shares; second is continuous R&D investment, approximately 216 million yuan, 204 million yuan, and 183 million yuan over the three years; third are marketing and market education costs that emerging industries must bear.

XREAL is not alone; Rokid has also experienced a long period of losses since its inception.

The industry generally agrees that smart glasses are an extremely capital-intensive track. Technologies such as ultra-compact optical engines, waveguide lenses, spatial perception algorithms, and high-density batteries all require massive upfront costs.

XREAL's decision to rush toward an IPO may partially stem from external shareholder pressure.

Public data shows XREAL has raised over 2.3 billion yuan in cumulative financing, with investors including Alibaba, Kuaishou, Sequoia China, and other prominent institutions. Founder Xu Chi holds approximately 27.98% of shares as the single largest shareholder, though this does not constitute a majority.

For a nine-year-old company with cumulative losses exceeding 2 billion yuan and cash reserves of less than 100 million yuan, an IPO is not only a means to raise funds but also the preferred path for investors to realize exits.

Rokid and Thunderbird face similar pressures.

Ai企查 information shows that in January 2026, Rokid completed a C+ round of financing, adding new shareholders such as Lens Technology, SenseTime, and Samsung Medical, with cash reserves exceeding 1.5 billion yuan.

Thunderbird secured over 1 billion yuan in financing from China Mobile and China Unicom earlier this year.

The "first AI glasses IPO" is likely to emerge within 2026; XREAL has filed its prospectus, while Rokid and Thunderbird, though not formally filing yet, remain under close market attention.

The current question is not whether they will go public, but who will be first.

图片

Meta's market cap is $1.7 trillion

Are China's AI glasses undervalued?

To ask the capital market for money, one must bring genuine collateral.

At present, the fundamentals of the AI glasses track appear solid.

First, while the market size is not large, growth potential remains significant, and no absolute leader has emerged yet. IDC predicts that global smart glasses shipments will exceed 23.687 million units in 2026; iiMedia Consulting data shows the global smart glasses market size was 12.58 billion USD in 2025 and is expected to reach 138.73 billion USD by 2029.

Second, the technological inflection point has arrived.

This year, AI glasses technology is branching from single-function to multiple forms, with features like recording, photography, translation, and payment being implemented in essential scenarios.

Additionally, tech giants are collectively betting big.

Over the past two months, manufacturers such as Rokid, Thunderbird, XREAL, Alibaba's Qianwen, and iFlytek have densely launched or previewed new products. Meta plans to double Ray-Ban Meta's annual production capacity to 20 million units by the end of 2026.

However, a "good track" does not equal a "good business."

The core contradiction currently facing AI glasses is that while market size is exploding, individual companies have yet to find a profitable model.

Industry organization statistics show that among the 200+ new AI features added to the industry, long-term user adoption rates are below 6%. Average return rates for AI glasses on major e-commerce platforms like JD.com and Tmall are as high as 30%.

图片

Chart: Meta Market Cap

If China's "first AI glasses IPO" needs a reference benchmark, it would be Meta in overseas markets, given Meta's dominant position abroad.

As of now, Meta's U.S. market cap stands at $1.7 trillion.

This $1.7 trillion corresponds to the following shipment volumes and market shares: Meta holds nearly 84% of the global non-display smart glasses market. Its 2025 AI glasses shipments reached 7.4 million units, a year-on-year increase of 281.3% (Counterpoint data), with the U.S. remaining the world's largest smart glasses market.

As for the Chinese market, according to Rokid founder Zhu Mingming's own sales target plan, Rokid aims to exceed 1 million units in 2026, reach 2 to 3 million units in 2027, and break through 10 million units in 2028.

In other words, it will take three years to catch up with the world's largest shipment volume, while where Meta will be in three years remains unknown.

Notably, the competitive landscape shows a stark contrast between the highly concentrated domestic market and overseas markets. In just one year, the top five market share list has undergone drastic changes.

This landscape means there are still variables in the domestic market, but it also means no enterprise can build a moat based on a single advantage.

Who can compete with Meta?

Based on the current competitive dynamics, domestic players can roughly be divided into four tiers.

图片

First Tier: Thunderbird Innovation, the "optical卷王" of the technology faction.

Thunderbird is one of the few manufacturers with full-chain self-research capabilities for core optical components and mass production capacity. Its advantage lies in the physical barriers of optical technology; the moat of upstream suppliers is physical limits, which are the most stubborn. However, Thunderbird's products are positioned at the mid-to-high end, making it difficult to reach the mass market, and its AI models rely on external supply.

图片

Second Tier: Rokid, the "diplomat" across ecosystems.

Rokid's most distinctive label is openness; its smart glasses can freely switch between mainstream large models such as Qianwen, DeepSeek, and Gemini. It has densely introduced supply chain enterprises as shareholders, turning suppliers into communities of shared interests. The brand awareness accumulated from its first-mover advantage and cooperation with CCTV's World Cup coverage are differentiation highlights, but profitability remains one of the key concerns for the market.

图片

Third Tier: Alibaba Qianwen, the strongest player in the ecosystem faction.

Qianwen AI glasses deeply integrate with ecosystems such as Alipay, Taobao, and Amap. This "hardware + AI + life services" trinity model is a typical approach for internet giants entering hardware. The certification of being "No. 1 in Omni-Channel Sales" from January to May 2026 proves the power of its ecosystem, but the challenge lies in hardware not being Alibaba's core competency; product definition and supply chain management still need time to refine.

图片

Fourth Tier: Xiaomi and Huawei, giants poised to launch.

Xiaomi leads China with a 28% shipment share, while Huawei's first HarmonyOS AI glasses start at 2,499 yuan. Both share common advantages: mature channel networks, enormous user bases, and deep supply chain management capabilities. Once AI glasses enter competition focused on "cost-effectiveness" and "ecosystem synergy," their potential momentum cannot be underestimated.

图片

As for iFlytek, with its 4,299 yuan pricing and positioning in 122-language translation, it mainly targets vertical scenarios in business translation, making it difficult to directly compete with the top four in the broad consumer AI glasses market for now.

Overall, the industry will likely remain in a "hundred glasses battle" chaotic stage for the next one to two years, with the ultimate winner determined by who survives last.

Rokid founder Zhu Mingming also stated at the 2026 Ecosystem and Developer Conference that "at this stage, the industry need not excessively pursue short-term market share; the entire track is still in its 'BlackBerry era,' and the critical point for true industrial-scale explosion has not yet arrived."

However, referencing the development trajectory of smartphones, in 2009 Nokia, BlackBerry, HTC, and Motorola rose together, but ultimately only a few players like Apple, Samsung, Huawei, and Xiaomi remained. AI glasses are likely to replay this script.

The maturation of upstream supply chains will lower entry barriers, but barriers related to brands, ecosystems, and channels will continue to rise.

In the end, no more than five companies may survive.

When AI glasses truly become the "next-generation computing platform," the essence of厂商 competition becomes the reallocation of model capabilities, terminal access, and ecosystem control.

Enterprises possessing large model capabilities, operating systems, hardware manufacturing capabilities, and channel coverage will have greater opportunities to prevail.

Note: Article images are from the internet; please contact for removal if there are copyright issues.

图片

图片

Momenta's Hong Kong IPO: What Justifies a 70 Billion Yuan Valuation?

2026 Coffee Franchisees: Walking Away and Demanding Refunds!

Three Hong Kong Stock Exchange Filings, Cumulative Losses Exceeding 800 Million Yuan: Can Jingze Bio's IPO Wait?

图片

暂无评论,快来发表第一条评论吧!No comments yet. Be the first!

📮 需求咨询📮 Inquiry