Sou Hu Cai Jing·2026-02-23 11:07Core Viewpoint - The capital market has shown strong support for major AI players like Tencent and Alibaba, with their stock prices rising significantly after the Lunar New Year, reflecting a positive response to the rapid deployment of large models in the past month [1][2]. Group 1: Stock Performance - On February 23, Tencent's stock closed at 538 HKD, up 3.07% from the previous day, while Alibaba's stock closed at 152.2 HKD, up 3.47% [2][6]. - Other AI companies also saw stock fluctuations, with Zhiyu's stock rising 42.7% to 725 HKD and MimiMax's stock increasing 14.5% to 970 HKD before experiencing a decline [6][7]. Group 2: AI Model Developments - The focus has shifted from AI performances at events to the practical applications of large models in everyday life, indicating a maturation of AI infrastructure [1]. - Zhiyu publicly addressed issues with its GLM-5 model and provided solutions, while the commercial success of Moon's Dark Side was highlighted, with Kimi's K2.5 model generating significant revenue shortly after its release [7]. Group 3: User Engagement and Activities - Tencent's Yuanbao platform reported over 36 billion lottery draws during the Spring Festival, with users completing over 10 billion AI tasks [18][19]. - The "Qianwen" app launched a promotional campaign that attracted over 1.3 million new users to AI ordering during the Spring Festival, indicating strong user engagement [11]. Group 4: Competitive Landscape - ByteDance is reportedly developing an AI chip in collaboration with Samsung, expanding its AI capabilities with a team of over 1,000 people [12]. - The competition among AI companies intensified during the Spring Festival, with various firms showcasing their models and capabilities, including the successful integration of AI in the Spring Festival Gala [14][15]. Group 5: Market Insights - The rapid user interactions and AI creations during the Spring Festival provided valuable real-world data for AI models, marking a significant milestone in user engagement and application [22][23]. - Analysts noted that the diverse and high-frequency user demands during the holiday period presented a unique opportunity for AI companies to gather training samples from actual user behavior [22].