来源：数字生命卡兹克 今天，智谱AI终于要上市了。 | 股票代码HK2513。 | | --- | | 从ChatGPT面世，到今天，过去了3年1个月零9天，一共，1136天。 | | 而从2025年4月14日智谱AI向证监会提交上市辅导备案，到今天正式港股上市，一共过去了269天。 | 真的蛮感慨的。 第一次去智谱参观，跟他们见面，还是2023年的4月。 时间真的眨眼而过。 好梦幻的三年。 作为中国大模型第一股，作为曾经AI六小龙里，最闪耀的新星，我想，对智谱AI你也有一定有很多好奇。 在今天这个时刻，我想，也用这一篇文章，6个信息。 帮你了解关于智谱AI的一切。 话不多说，我们，开始。 一. 智谱AI是谁？ 智谱 AI，全称"北京智谱华章科技股份有限公司"。 2019年6月11日成立，出身非常标准的清华系。 公司技术根基来自清华大学计算机系知识工程实验室，也就是KEG Lab。 这个实验室十几年一直在做知识图谱、NLP、多模态等底层研究，最早做出来的东西之一是学术搜索平台AMiner。 CEO张鹏算是学术出身+创业型的典型，原来就在KEG做AMiner，2019年出来把团队改造升级成智谱，把很多的科研成果 ...

今天，智谱AI终于要上市了。 | | | 真的蛮感慨的。 第一次去智谱参观，跟他们见面，还是2023年的4月。 公司技术根基来自清华大学计算机系知识工程实验室，也就是KEG Lab。 时间真的眨眼而过。 好梦幻的三年。 作为中国大模型第一股，作为曾经AI六小龙里，最闪耀的新星，我想，对智谱AI你也有一定有很多好奇。 在今天这个时刻，我想，也用这一篇文章，6个信息。 帮你了解关于智谱AI的一切。 话不多说，我们，开始。 一. 智谱AI是谁？ 智谱 AI，全称"北京智谱华章科技股份有限公司"。 2019年6月11日成立，出身非常标准的清华系。 这个实验室十几年一直在做知识图谱、NLP、多模态等底层研究，最早做出来的东西之一是学术搜索平台AMiner。 CEO张鹏算是学术出身+创业型的典型，原来就在KEG做AMiner，2019年出来把团队改造升级成智谱，把很多的科研成果，往商业公司方向重构。 而智谱的灵魂人物，首席科学家唐杰老师，之前就是实验室的领头人。 实验室里除了智谱的人之外，还有一个AI圈的非常有名的创业大佬，跟KEG有千丝万缕的关系，名字我就不提了，但是你们肯定都知道。 所以，说KEG Lab是中国AI的 ...

Core Insights - During the "14th Five-Year Plan" period, China's express delivery business volume reached a trillion level, growing to nearly 200 billion packages, contributing over 60% to global growth [1] - The postal industry's revenue increased to 1.8 trillion yuan, with an average annual growth rate exceeding 10%, ranking among the top in modern service industries [1] - The per capita annual express delivery usage rose to 141 packages, with a peak daily volume of 777 million packages [1] Industry Development - The breadth and depth of postal services in rural and border areas have significantly improved, with the number of operating outlets increasing by nearly 1.5 times compared to five years ago [1] - All border natural villages in the country have achieved postal service access, and regions like Xinjiang and Tibet have ended the history of e-commerce packages not being "free shipping" [1] - Over the past five years, the share of express delivery revenue and volume from central and western regions has increased by 5.4 and 7.5 percentage points, respectively, facilitating better integration of these regions into the national unified market [1] Technological Advancements - The service capabilities of the postal industry have continuously improved, with sorting centers achieving near-automation and the scale of unmanned warehouse operations expanding [1] - Pilot programs for unmanned vehicles have shown positive results, and drones have achieved regular operations in some areas [1] - The application of cloud computing and large models in demand forecasting, intelligent warehouse management, intelligent scheduling, and route planning has accelerated [1] - The comprehensive transportation efficiency, informatization, organizational level, service range, and cost-effectiveness of postal and express services remain among the best in the world [1]

1月7日晚间，智谱AI在港交所公告发售价及配发结果，香港公开发售获1159.46倍认购，国际发售获15.28倍认购；发售价116.20港元，所得款项净 额41.734亿港元。今日智谱港股暗盘表现亮眼，一度涨近38%至160港元。 智谱将于1月8日开始在香港联交所主板买卖，成为全球资本市场首家以通用人工智能（AGI）基座模型为核心业务的上市公司。 早在2025年4月，智谱曾在证监会北京监管局开启A股上市辅导备案。但截至12月12日，公司并未收到中国证监会关于推进A股上市的进一步意见 或问询。在此背景下，智谱选择转向港股，为这场高投入、长周期的大模型竞赛寻找更可持续的燃料。 | 有效申請數目 | 204,871 | | --- | --- | | 獲接納申請數目 | 50.907 | | 認購水平 | 1,159.46倍 | | 觸發回補機制 | 是 | | 公開發售項下初步可供認購發售股份數目 | 1,871,000 | | 公開發售項下發售股份的最終數目(重新分配後) | 7,483,900 | | 公開發售項下發售股份數目佔全球發售項下發售股份數目的百 20.00% | | | 分比 | | GLM技术体系 ...

Investment Rating - The report does not provide a specific investment rating for the company [1]. Core Insights - The company, Zhipu AI, is a leading independent general large model developer in China, established in 2019, and has developed a unique General Language Model (GLM) framework that excels in long text understanding and logical reasoning [6][12]. - Zhipu AI's flagship products, GLM-4.5 and GLM-4.7, have achieved top rankings in international benchmark tests and have gained significant recognition in the global developer community [6][14]. - The company has a strong market position, ranking first among independent general large model developers in China with a market share of 6.6% as of 2024 [6][15]. - Zhipu AI plans to go public on the Hong Kong Stock Exchange on January 8, 2026, with an IPO price of HKD 116.20 per share, aiming to raise approximately HKD 4.3 billion [6][15]. Summary by Sections 1. Company Overview - Zhipu AI is built on technology from Tsinghua University and aims to compete with OpenAI, focusing on a self-regressive fill-in-the-blank GLM framework [12][16]. - The company has released several significant models, including the GLM-130B, which marked its entry into the mainstream large language model market [12][16]. 2. Business Model and Operations - The business model is centered around Model as a Service (MaaS), offering both localized and cloud deployment options [21][24]. - Localized deployment accounts for a significant portion of revenue, with high margins, while cloud deployment is rapidly growing and aims to capture a larger market share [24][25]. 3. Historical Financial Analysis - Revenue has shown rapid growth, with projections indicating revenues of CNY 785 million in 2025 and CNY 1.55 billion in 2026, reflecting a compound annual growth rate of over 130% from 2022 to 2024 [1][30]. - The company has been operating at a loss due to substantial R&D investments, with cumulative R&D expenses exceeding CNY 4.4 billion from 2022 to 2024 [6][30]. 4. Core Competitiveness - Zhipu AI's competitive edge lies in its fully self-developed technology system, leading model performance, and a robust open-source ecosystem [38][39]. - The GLM series models have demonstrated significant advantages in various applications, including multi-modal understanding and generation [39][40]. 5. Profitability Forecast and Investment Suggestions - The company is expected to achieve revenues of CNY 7.9 billion in 2025, CNY 15.5 billion in 2026, and CNY 32.2 billion in 2027, with a gradual shift towards cloud-driven revenue [6][7]. - The overall gross margin is projected to reach 50% by 2025, with improvements in cloud margins as the business scales [6][7].

Core Insights - Xiaomi's transformation by 2025 is driven by three key factors: the density of its engineering team, long-term R&D investment capability, and the will to continue betting on uncertain directions [1][53][102] - The focus is not merely on the success of individual products but on the formation of a cohesive structure within the larger AI wave, where the competition is about having complex, real, and continuous application scenarios [1][53] Smartphone Business - Xiaomi's smartphone shipments in Q3 2025 reached 43.4 million units, maintaining a global market share of 13.6%, and showing a 1.3% year-on-year increase [3][54] - In the Chinese market, Xiaomi's sales in the 4000 yuan and above price segment rose to about 25%, indicating a shift from occasional high-end hits to a structural component of its offerings [5][56] - The smartphone has evolved from a high-cost-performance terminal to an ecological entry point, serving as a personal identity credential and a control hub for the interconnected ecosystem of vehicles and home appliances [5][56] Automotive Business - Xiaomi's electric vehicle deliveries reached a record high of 108,800 units in Q3 2025, with total deliveries exceeding 250,000 units in the first three quarters, and an annual target of over 400,000 units [8][59] - The automotive business transitioned from loss-making to operational profitability, with Q3 2025 revenue of approximately 28.3 billion yuan, nearly tripling year-on-year, and a gross margin of around 25% [11][62] - Xiaomi's self-developed intelligent driving technology has received multiple high-level driving licenses, showcasing its commitment to full-stack self-research in smart driving [13][64] AIoT Ecosystem - By Q3 2025, Xiaomi's AIoT platform connected over 1 billion devices, marking a significant increase from 624 million devices at the end of 2023 [22][73] - The AIoT ecosystem has evolved from "connecting everything" to "intelligent interconnection," with the voice assistant, Xiao Ai, upgraded to serve as a smart home hub capable of executing complex cross-device commands [23][74] - Xiaomi's AIoT devices are now part of a dynamic system that integrates user intent, cloud-based models, and home devices, enhancing the overall user experience [22][73] R&D and Manufacturing - Xiaomi's R&D investment in 2025 is projected to exceed 30 billion yuan, with a focus on building long-term capabilities rather than just optimizing functions and experiences [34][96] - The establishment of the Wuhan smart home appliance factory, with a planned annual capacity of 7 million air conditioners, represents a significant step in achieving a complete closed-loop from R&D to mass production [18][69] - Manufacturing capabilities have been enhanced through automation and digitalization, with the Beijing smartphone factory achieving high production efficiency and the automotive factory implementing advanced manufacturing techniques [41][94] Future Outlook - Xiaomi's strategy is to integrate all dispersed devices, scenarios, data, and capabilities into a cohesive system, positioning itself uniquely in the AI landscape [99][101] - The company is seen as an "ecological empire," with a complete structure that connects smartphones, vehicles, home appliances, and IoT, continuously expanding its boundaries and capabilities [50][101] - Xiaomi's long-term vision emphasizes patience in building foundational capabilities, which will support the expansion of its interconnected ecosystem [102]

Core Viewpoint - MiniMax, an artificial general intelligence company, successfully completed its IPO in Hong Kong, achieving a record oversubscription rate of 1209 times, reflecting strong investor interest in the AI sector [1]. Group 1: IPO Details - MiniMax's IPO involved a total of 25.39 million shares, with a price range set between HKD 151 and HKD 165, aiming to raise approximately HKD 38.34 billion to HKD 41.89 billion [1]. - The IPO attracted 14 cornerstone investors, including notable institutions such as Aspex, Eastspring, and Mirae Asset, with a total subscription amount of HKD 27.23 billion, accounting for 69.8% of the global offering [1]. - The company plans to allocate about 90% of the raised funds for research and development over the next five years, with 70% dedicated to developing large models and 20% for AI-native products [1]. Group 2: Company Overview - Founded in 2022, MiniMax focuses on advanced large model research and AI-native product development, continuously iterating its technology to enhance intelligence levels [2]. - The company has developed a diverse range of AI-native products, including MiniMax, Hailuo-02, and Speech-02, which support multimodal understanding and generation across text, visuals, and audio [2]. - MiniMax ranks as the tenth largest model technology company globally, with a market share of 0.3%, despite being a relatively resource-limited startup competing against larger, well-resourced firms [2]. Group 3: Industry Context - The strong demand for MiniMax's IPO reflects a broader trend in the Hong Kong market, where other AI companies, such as Zhiyuan AI and Birun Technology, have also experienced significant oversubscription rates of 1164 times and 2348 times, respectively [4].

每经记者｜可杨 杨卉 每经编辑｜何小桃 魏文艺 2026年国际消费电子展（CES 2026）开幕首日，英伟达CEO黄仁勋在离主场馆不远的酒店搭起了自己的"台子"。 美国当地时间1月6日中午，前一日刚刚结束CES演讲的黄仁勋在拉斯维加斯举办了一场媒体沟通会，与包括《每日经济新闻》记者在内的媒体面对面交 流。 在近一个半小时的交流中，黄仁勋始终保持着他的"招牌姿势"：坐在椅子边缘，一条腿撑地，时不时转身面向提问记者。在前一日演讲的基础上，黄仁勋 此次还提到了H200芯片、中国市场、汽车、大模型、人工智能、GPU（图形处理单元）、机器人等热门议题，还笑称自己可以接受"亿万富豪税"。 黄仁勋带领英伟达冲在AI（人工智能）的潮头，不仅坐拥巨量财富，也在科技产业界有着巨大影响力。他在回答关于自己是任职时间最长的CEO这一问 题时再次展现出独特的幽默感："秘诀是第一不要被解雇，然后第二，不要觉得无聊。" 不过，采访过程中也有部分媒体吐槽此次采访内容过于"温和"，黄仁勋并未直面回答较为犀利的问题。 谈中国创业公司：中国工程师充满创业精神 黄仁勋：我认为，中国涌现出的大量初创企业，以及其中许多已经上市并表现得非常出色的公司，都 ...

（文章来源：每日经济新闻） 每经AI快讯，1月7日，国产大模型公司智谱港股暗盘涨近20%，报138港元。 ...

Core Insights - The Ministry of Industry and Information Technology (MIIT) has issued the "Action Plan for the Integration of Industrial Internet and Artificial Intelligence" aimed at enhancing the intelligence level of industrial internet platforms [1] Group 1: Key Initiatives - The plan emphasizes the use of deep learning and large models to strengthen capabilities in element connectivity, intelligent analysis, and resource allocation within industrial internet platforms [1] - It encourages industrial internet platform companies to accelerate innovation in low-code and no-code technologies based on artificial intelligence, thereby improving the efficiency of industrial APP development and system integration [1] - The initiative explores the creation of a "model pool" relying on industrial internet platforms, aiming to cultivate and launch a batch of industrial model products [1] Group 2: Technological Development - The plan promotes the compatibility of underlying architecture, data protocols, and artificial intelligence, fostering innovation in domestic intelligent agent standard protocols [1] - It aims to develop innovative models such as "industrial internet platform + intelligent agents," focusing on typical scenarios like production network optimization, human-machine interaction, intelligent equipment health management, and supply chain optimization [1]