记者丨张伟泽 编辑丨李莹亮,和佳 继昨日大涨后，今日港股AI股开始回调走弱。1月13日午后，港股AI应用板块午后继续降温。截至发 稿，智谱(02513.HK)跌超13%，MINIMAX-WP(00100.HK)跌10%，商汤(00020.HK)跌超4%，港股科网 股多股下挫，快手、华虹半导体、金山软件跌超2%，阿里巴巴、哔哩哔哩、小鹏汽车飘红，截至发稿 涨幅超2%。 | 名称 | 现价 | 涨跌 | 涨跌幅 - | | --- | --- | --- | --- | | 商汤-W | 2.530 | -0.110 | -4.17% | | 快手-W | 77.850 | -2.400 | -2.99% | | 花机未曾体 | 89.550 | -2.000 | -2.18% | | 金山软件 | 31.280 | -0.660 | -2.07% | | 小米集团-W | 38.040 | -0.700 | -1.81% | | 中本国际 | 74.000 | -1.300 | -1.73% | | 百度集团-SW | 142.600 | -2.100 | -1.45% | | 美的集团 | 85.200 | ...

记者丨 张伟泽 编辑丨李莹亮,和佳 继昨日大涨后，今日港股AI股开始回调走弱。1月13日午后，港股AI应用板块午后继续降温。截至发稿， 智谱(02513.HK)跌超13%， MINIMAX-WP(00100.HK)跌10%，商汤(00020.HK)跌超4%，港股科网股多股下挫，快手、华虹半导体、金山软件跌超2%，阿里巴巴、哔哩 哔哩、小鹏汽车飘红，截至发稿涨幅超2%。 | 截至1月13日14：17走势 | | --- | 目前港股AI产业版图正在不断拓展。在AI产业链的上游，有中芯国际、壁仞科技、天数智芯等多家提供算力和数据支撑的硬件服务企业。而在 AI产业链的中游，除了刚刚上市的MiniMax和智谱之外，也有阿里、腾讯和百度等参与AI软件的综合厂商。在AI产业链的下游，则有百度旗下 的AI芯片公司昆仑芯，以及存储芯片公司芯天下技术等企业，均已向港交所提交上市申请。 这股热潮并非偶然。瑞银全球投资银行亚太股票资本市场联席主管胡凌寒在近期接受21世纪经济报道记者专访时指出，AI是当前全球范围内热 度极高的赛道，随着行业发展，这些公司必然产生融资需求，可以预见2026年一定会有更多AI产业链相关的公司登陆香港市场 ...

每经AI快讯，1月13日，港股智谱午后持续下挫跌超13%，MINIMAX跌超8%。 ...

Group 1 - The Hong Kong stock market's AI application sector experienced a decline in the afternoon, with notable drops in several companies [1] - Zhihui (02513.HK) fell over 10%, while MINIMAX-WP (00100.HK) decreased by nearly 9% [1] - Other companies such as SenseTime (00020.HK) and Fourth Paradigm (06682.HK) also saw declines [1]

Core Insights - The AI industry is accelerating rapidly, marked by the IPOs of companies like Zhiyu and MiniMax, indicating a significant milestone in the global capital market with the first AGI-based model company going public [1][3][7] - Zhiyu's annual recurring revenue (ARR) for its MaaS product saw a staggering growth of over 900% in 2025, outpacing many leading overseas AI companies [5][9] - MiniMax's IPO on January 9, 2026, showcased an extraordinary market response, with its stock price surging over 80% at one point, leading to a market capitalization exceeding 900 billion HKD [6][7] Company Developments - Zhiyu's founder and chief scientist, Tang Jie, announced the upcoming release of GLM-5, signaling continued innovation and growth in their product offerings [3] - The rapid growth trajectory of both Zhiyu and MiniMax reflects a broader trend in the AI sector, where companies are achieving significant milestones in record time, with MiniMax's journey from inception to IPO taking just four years [8][9] Industry Trends - The AI sector is transitioning from a technological trend to a phase of accelerated capital, commercial, and engineering integration, characterized by leapfrog advancements in model capabilities and increased complexity in AI systems [9][10] - The testing and development landscape is evolving, necessitating that testing professionals adapt to new methodologies that account for AI's unpredictable behavior and the complexities of AI systems [11][12] - There is a growing demand for professionals who can effectively test AI systems, with many individuals seeking to enhance their skills to ensure they can adequately assess AI outputs and integrate testing capabilities into existing engineering frameworks [12][13]

Group 1 - The stock of Zhizhu experienced significant volatility, dropping nearly 10% to reach 188 HKD after previously rising over 8% to a price of 225.2 HKD. As of the report, Zhizhu's decline was over 8.01% with a trading volume exceeding 400 million HKD [1][7]. Group 2 - China Galaxy Securities predicts that by 2026, AI applications will transition from "AI empowerment" to "AI native." Companies that can internalize AI capabilities into their product architecture will gain a competitive advantage in the upcoming market [3][9].

人民财讯1月13日电，截至发稿，港股智谱震荡走低，现跌超6%。 （原标题：港股智谱震荡走低，现跌超6%） ...

Group 1 - The core viewpoint of the news is the strategic cooperation between Zhipu, referred to as the "first stock of global large models," and Didi, focusing on the exploration of key technologies in general artificial intelligence (AGI) and their application in the transportation sector [1][2] - Didi has been increasing its investment in large models and intelligent agents, developing innovative applications such as AI travel assistants and business travel assistants based on real and complex transportation scenarios [1] - Zhipu has a strong foundation in large model architecture, training paradigms, and intelligent agent technology, and the collaboration aims to advance the implementation of agent scenarios and talent cultivation in the large model field [1] Group 2 - Didi's representative expressed expectations for both parties to leverage their respective strengths to promote the application and development of general artificial intelligence in broader real-world scenarios [2] - Zhipu was listed on the Hong Kong Stock Exchange on January 8, with an initial share price of HKD 116.20, and as of the report, the share price rose to HKD 219.8, resulting in a total market capitalization of HKD 96.763 billion (approximately RMB 86.725 billion) [4]

Group 1 - The core viewpoint of the news is that Zhiyun (02513.HK) has formed a strategic partnership with Didi to explore key technologies in General Artificial Intelligence (AGI) and their applications in the transportation sector [1] - Didi has been increasing its investment in large models and intelligent agents, leading to innovations such as AI travel assistants and business travel assistants [1] - Zhiyun has a strong foundation in large model architecture, training paradigms, and intelligent agent technology, and the partnership aims to enhance the deployment of agents in complex business scenarios [1] Group 2 - Zhiyun was established in 2019 and focuses on developing advanced general large models, launching China's first proprietary pre-trained large model framework, GLM, in 2021 [2] - The company has achieved significant growth in its cloud-based Model as a Service (MaaS) and subscription business, with over 2.9 million users on its API platform [2] - According to Frost & Sullivan, Zhiyun ranks first among independent general large model developers in China and second overall, with a market share of 6.6% as of 2024 [2] Group 3 - Zhiyun's R&D investments from 2022 to 2024 are projected to be 84 million yuan, 529 million yuan, and 2.195 billion yuan, with 1.595 billion yuan allocated for the first half of 2025 [3] - The company has a research team of 657 members, with R&D personnel making up 74% of its workforce [3] - On January 12, Zhiyun's stock surged by 31.40%, closing at 208.4 HKD per share, with a market capitalization of 91.744 billion HKD, reflecting a cumulative increase of 79.35% since its IPO [3]

Core Viewpoint - The listing of Zhipu, referred to as the "first global large model stock," marks a significant moment in the AI industry, highlighting both the potential and challenges faced by AI companies in achieving profitability and sustainable growth [1][3][25]. Company Overview - Zhipu officially listed on the Hong Kong Stock Exchange with an opening price of 124.1 HKD, reflecting a 6.88% increase from its issue price [3]. - Founded in 2019 from Tsinghua University's Knowledge Engineering Laboratory, Zhipu has raised a total of 8.3 billion RMB from various investors, including major firms like Hillhouse Capital and Tencent [5]. - Despite significant funding, Zhipu reported a loss of 2.958 billion RMB in 2024 and an additional 2.358 billion RMB in the first half of 2025, accumulating losses exceeding 5.3 billion RMB since its inception [5]. Financial Performance - Revenue has been increasing, with a compound annual growth rate (CAGR) of 130% from 2022 to 2024, and a 325% year-on-year increase in revenue to 191 million RMB in the first half of 2025 [7]. - The revenue growth, however, is minimal compared to the substantial losses, indicating a significant imbalance between income and expenditure [7]. Team Structure and Strategy - Zhipu operates under a collective decision-making model led by three key figures: Chief Scientist Tang Jie, who developed the GLM model framework, Zhang Peng, who oversees commercialization, and Liu Debing, responsible for technical implementation [9]. - The company has strategically focused on B2B applications rather than consumer-facing products, with 84.8% of its revenue in the first half of 2025 coming from local deployment orders [11]. Market Position and Competition - The competitive landscape for large model companies has narrowed, with only a few players like Zhipu, Moonlight Dark Side, and MiniMax remaining prominent [16]. - Zhipu plans to raise 4.3 billion HKD from its IPO, with 70% allocated to the development of general AI models, including the upcoming GLM-5 model [18]. Future Outlook - The company aims to target emerging markets, particularly along the Belt and Road Initiative and Southeast Asia, to establish an international collaborative alliance for autonomous large models [20]. - Challenges include resource constraints, such as the availability of A100 chips and the high cost of talent, with senior algorithm engineers commanding salaries around 2 million RMB [21]. - The CEO's statement about racing against time rather than competitors reflects the urgency in the AI sector, as the anticipated "explosion year" for AI replacement approaches in 2026 [23]. Industry Implications - The listing of Zhipu signifies a new phase in the capital dynamics of the AI industry, revealing the tension between technological aspirations and commercial realities [25]. - The ongoing challenge for AI entrepreneurs is to balance idealism with practical business strategies, as the industry moves towards a more sustainable model [27].