机构：华泰证券 研究员：夏路路/侯杰 据百度公司1 月2 日公告，其已根据上市规则PN15 向香港联交所提交有关建议分拆的方案，而香港联 交所已确认可进行建议分拆。我们认为昆仑芯作为国产AI 计算芯片龙头之一，未来其若顺利实现分拆 上市，则上市后更加清晰的信息披露有望助力市场充分认知该业务的内在价值，并推动该业务分部在百 度集团估值中实现进一步的价值重估。展望后续，我们看好在昆仑芯、自动驾驶等子业务线向好发展与 潜在资本市场计划的助推下，百度集团获得相应的价值重估机遇。与此同时，百度集团丰厚的净现金储 备为其估值提供了一定的支撑，后续若公司加大股东回报力度，则有望进一步加强其估值吸引力。建议 持续关注百度各AI 业务项的进展及资本市场计划的推进节奏。 我们认为在AI 时代下，百度集团丰富多元的业务线具备全新发展机遇，后续伴随业务的放量及基本面 的逐步改善，有望进一步推动百度集团的价值重估。1）自动驾驶：12 月22 日，萝卜快跑宣布将于2026 年率先在英国伦敦开启无人驾驶测试及出行服务，是中国无人驾驶首次驶入英国伦敦，进入中国香港以 外的右舵市场（据Wind 报道）。2）AI agent：12 月11 日， ...

Core Insights - The article highlights the rapid development and adoption of AI applications in China, marking a shift from mere technological showcases to practical productivity tools across various industries [1][2][3] - Baidu is positioned as a leading player in this AI application landscape, having established a comprehensive system from chips to models to applications, reflecting the broader trajectory of China's AI development [1][2][24] Group 1: AI Application Trends - The global tech community has recognized the innovation in Chinese AI applications, with significant contributions from companies like Baidu and DeepSeek, as evidenced by 22 out of the top 50 AI mobile applications being from China [2] - The differentiation in AI strategies between the US and China is becoming more pronounced, with US companies focusing on model performance while Chinese firms are rapidly deploying diverse applications [2][22] Group 2: Baidu's AI Strategy - Baidu's AI strategy is characterized by a systematic approach, integrating various components of AI technology, which serves as a microcosm of China's overall AI path [1][24] - The launch of the "秒哒" (MiaoDa) no-code application development platform has enabled non-programmers to create AI applications, demonstrating the democratization of AI technology [7][15] Group 3: Economic Impact and Market Response - The economic value generated by AI applications through platforms like 秒哒 has exceeded 5 billion yuan, showcasing the tangible benefits of AI in various sectors [15] - Baidu's stock has seen a positive response in the capital markets, reflecting investor confidence in the company's AI-driven growth potential [3][25] Group 4: Case Studies of AI Applications - The "蜗牛牙牙" (Snail Tooth) application, developed for children's oral health, exemplifies how AI can meet real user needs without requiring complex technical skills [10][11] - The "油气井生产优化设计系统" (Oil and Gas Well Production Optimization System) developed using 秒哒 illustrates how industry professionals can leverage AI tools to address specific operational challenges [11][14] Group 5: Future Prospects and Innovations - Baidu's "伐谋" (FaMou) intelligent agent represents a significant advancement in AI capabilities, enabling self-evolution and optimization in complex industrial scenarios [17][19] - The upcoming launch of new Kunlun chips is expected to enhance Baidu's AI infrastructure, further solidifying its competitive edge in the AI landscape [25][26] Group 6: Global Expansion and Recognition - Baidu's AI applications are not only gaining traction domestically but are also expanding into international markets, with products like 秒哒 and 慧播星 (HuiBoXing) achieving recognition abroad [27] - The comprehensive self-research capabilities across various AI layers position Baidu uniquely in the global tech landscape, allowing it to drive scalable AI applications [24][26]

Core Insights - The AI progress in 2025 is characterized by a significant shift, indicating a transition from technology to a powerful force reshaping the global economy, similar to the internet and mobile internet eras [2][19] - Nvidia has become the world's first company to reach a market capitalization of $5 trillion, symbolizing a pivotal moment in the valuation of computing power as a strategic asset [4] - The global consumption of tokens has increased nearly tenfold in two years, with Baidu's AI search monthly active users reaching 382 million, indicating a shift in information access from human-driven to model-driven interactions [7] Industry Trends - AI infrastructure is experiencing unprecedented growth, with Microsoft, Amazon, and Google investing $200 billion over two years to expand computing power and data centers, marking the largest infrastructure project in digital economic history [10] - The Nasdaq Top 20 semiconductor companies are gaining weight, reflecting a shift in the global value chain from "connecting the internet" to "building intelligence" [10] - The AI search sector is rapidly evolving, with Baidu leading in monthly active users, as consumers increasingly prefer AI-driven search for making purchasing decisions [17] Financial Performance - The revenue growth rates for AI-related companies in the US show a significant upward trend, with projections indicating continued growth into 2025 [12] - The net profit growth rates for these companies are also expected to rise, reflecting the overall positive outlook for the AI sector [12] Application Development - The year 2025 marks the official start of the AI Agent era, with a notable increase in demand for large model tools and applications across various companies [14] - Major players like Microsoft and Google are actively developing AI agents, indicating a competitive landscape in AI application deployment [14][15]

最具象征意义的事件来自市场：英伟达跃升为全球首个市值5万亿美元的公司。这不仅是估值的突破，更是世界对未来生产范式的投票——算力正成为一种 新的战略资产，取代移动互联网时代的流量与终端。 "透视图"栏目在年终特别策划了"36氪年度总结"系列，用数据透视2025全年趋势，以图片呈现今年商业世界中不可错过的要点。 作者丨范亮 2025年的AI进程呈现出一种罕见的清晰度：它不再是一项技术，不再是一条赛道，而是一种力量迁移，就像当初的互联网、移动互联网时代一样。资本、 计算与需求的方向同时发生偏转，这种一致性通常只在全球经济结构重排的时期出现。 36氪制图 36氪制图 使用侧同样出现了罕见加速度。全球Token消耗在两年内增长近十倍；在中国，百度AI搜索月活达到3.82亿，首条结果的富媒体覆盖率已经达到70%。这意 味着信息入口的主权开始从人转向模型，从输入式交互转向委托式代理。 36氪制图 尽管外界对于"AI泡沫"的讨论仍在继续，但从技术扩散规律看，争议本身就是新周期的一部分，而非趋势的反证。价格可能波动，但资本开支、供给能力与 用户行为的变化更具决定性。 AI基建正迎来史无前例的高潮。微软、亚马逊、谷歌两年投入20 ...

Core Insights - Alibaba's CEO, Wu Yongming, emphasizes the necessity for an AI-native consumer-facing super entrance, leading to the launch of the "Qianwen" app, which achieved over 10 million downloads within a week of public testing, making it the fastest-growing AI application to date [2][4] - The AI-to-consumer market is still open for latecomers, particularly for productivity-oriented AI, despite earlier entrants like ByteDance's Doubao and DeepSeek establishing significant user bases [2][5] - Alibaba's strategic investment in AI and cloud infrastructure reached 120 billion yuan over the past four quarters, indicating a strong commitment to AI development despite short-term profit pressures [2][10] Group 1: Market Dynamics - The AI market is becoming increasingly competitive, with major players like Alibaba, Ant Group, and others launching new AI applications to capture consumer interest [4][5] - Current market leaders include ByteDance's Doubao with 172 million monthly active users, DeepSeek with 145 million, and Tencent's Yuanbao with 32.86 million, while Alibaba's previous app had only 3.06 million [4] - Wu Yongming's AGI three-stage theory outlines the evolution of AI capabilities, indicating that the market is transitioning from basic language imitation to more complex task execution [4][5] Group 2: Product Features and User Engagement - Both "Qianwen" and Ant Group's "Lingguang" emphasize usability, with features designed to simplify complex tasks, such as generating applications from natural language in 30 seconds [5][6] - The future vision for "Qianwen" includes integration with various life scenarios like maps, food delivery, ticket booking, and shopping, aiming to create a comprehensive AI life entrance [6][9] Group 3: Infrastructure and Ecosystem - The competition for consumer entry points is not just about products but also about underlying infrastructure and ecosystem capabilities, with Alibaba's full-stack AI strategy being a significant advantage [8][9] - Alibaba has open-sourced over 300 AI models, with total downloads exceeding 600 million, establishing a strong presence in the global AI landscape [8][9] - The company plans to invest over 380 billion yuan in cloud and AI hardware infrastructure over the next three years, indicating a long-term commitment to enhancing its capabilities [9][10] Group 4: Financial Implications - Alibaba's recent financial reports show a significant decline in profits, with a 85% year-on-year drop in operating profit, attributed to heavy investments in AI and cloud infrastructure [10][11] - The CFO noted that while core business revenues are strong, the focus on AI commercialization will lead to fluctuations in short-term profitability [10][11] - Wu Yongming expressed confidence that an AI bubble is unlikely in the next three years, as demand for AI resources is expected to exceed supply [10][11]

Core Insights - Alibaba's CEO, Wu Yongming, emphasizes the necessity for an AI-native consumer-facing super entry point, leading to the launch of the "Qianwen" app [1][2] - The app achieved over 10 million downloads within a week of public testing, surpassing the growth rates of competitors like ChatGPT and DeepSeek, marking it as the fastest-growing AI application to date [3][5] - Despite facing short-term challenges such as profit pressure and technology integration, Alibaba's strategic focus on AI indicates a commitment to not miss out on the AI ecosystem reshaping [3][18] Market Dynamics - The AI market is becoming increasingly competitive, with major players like ByteDance, Tencent, and Baidu already establishing a strong presence in the consumer AI sector [6][8] - According to QuestMobile, the current market is dominated by three major players: ByteDance's Doubao with 172 million monthly active users, DeepSeek with 145 million, and Tencent's Yuanbao with 32.86 million [6] Strategic Positioning - Wu Yongming's AGI three-stage theory underpins Alibaba's push into the consumer AI market, identifying the current phase as "assisting humans" [7][8] - The focus is shifting from entertainment-based AI to productivity-oriented AI, which is seen as the true growth area for consumer applications [8] Future Vision - The "Qianwen" app aims to integrate various life scenarios such as maps, food delivery, ticket booking, and shopping, positioning itself as a future AI life entry point [11][18] - The competition for consumer entry points is viewed as a battle for future traffic distribution rights, with successful AI applications expected to directly understand user needs and deliver services [11][12] Infrastructure and Ecosystem - Alibaba's comprehensive AI strategy includes significant investments in AI and cloud infrastructure, with a reported capital expenditure of 120 billion yuan over the past four quarters [3][15][19] - The company has open-sourced over 300 AI models, achieving over 600 million downloads, which positions it as a leading player in the global open-source AI landscape [14][15] Financial Implications - Alibaba's aggressive investment in AI has led to a significant decline in profits, with a reported 85% year-on-year drop in operating profit for Q2 of fiscal year 2026 [18][19] - The CFO noted that while core business revenues are strong, the focus remains on long-term growth through AI investments, despite short-term profitability fluctuations [19]

Core Insights - Baidu's AI business shows strong growth, with total revenue reaching 96 billion RMB, a year-on-year increase of over 50%, becoming a key driver for long-term growth amid traditional business pressures [1][3] Financial Performance - Baidu reported total revenue of 312 billion RMB for Q3 2025, a year-on-year decline of 7%. The net profit attributable to Baidu under non-GAAP was 37.7 billion RMB, down 36% [2] - Core revenue was 247 billion RMB, also down 7% year-on-year, with non-GAAP net profit for core operations at 38 billion RMB, a decrease of 32% [2] AI Business Segments - The AI business consists of three categories, with revenue of 96 billion RMB, showing a year-on-year growth of over 50% [3] - AI infrastructure revenue reached 42 billion RMB, a 33% increase year-on-year, with high-performance computing subscription revenue growing by 128% [3] - AI applications generated 26 billion RMB in revenue, while AI-native marketing services achieved 28 billion RMB, a remarkable growth of 262% [3] User Engagement and Market Position - Baidu AI search has a monthly active user base of 3.82 billion, reflecting an 18.63% quarter-on-quarter increase, maintaining its leading position in the AI search industry [3] - The total monthly active users for Baidu's flagship AI applications, including Baidu Wenku and Baidu Wangpan, are nearing 300 million [3] Other Business Highlights - The "Luobo Kuaipao" service saw a significant increase in global ride service instances, reaching 3.1 million, a 212% year-on-year growth [4] - The service has expanded its global footprint, covering 22 cities and achieving over 1.4 billion kilometers in autonomous driving mileage [4] - Baidu's founder highlighted the transformative value of AI in business, emphasizing the robust growth of AI cloud and the rapid expansion of the "Luobo Kuaipao" service [5]

Core Viewpoint - Warren Buffett's decision to invest in Google and include it in Berkshire Hathaway's top ten holdings signals a shift in the value investment paradigm, acknowledging the long-term value of AI [1] Group 1: Investment in AI - Google has built a comprehensive AI ecosystem with its self-developed TPU chips, Gemini model, and extensive applications in search, YouTube, and cloud services, which aligns with Buffett's value investment principles [1] - Baidu is recognized as a leading player in the global AI landscape, with a complete stack of self-developed technologies across chips, frameworks, models, and applications, showcasing a significant competitive advantage [1][3] Group 2: Financial Performance - Baidu's AI business revenue surged by 50% year-on-year, with AI cloud revenue increasing by 33%, and AI native marketing service revenue skyrocketing by 262% to 2.8 billion [1][2] - Baidu's total investment in AI has exceeded 100 billion since March 2023, indicating a strong commitment to maintaining its technological leadership in the AI sector [2] Group 3: Business Transformation - Baidu is undergoing a significant transformation by integrating AI into its entire product line, which is expected to drive substantial revenue growth [2][15] - The company has restructured its search product, with 70% of mobile search results now featuring AI-generated content, leading to a notable increase in active users [4][5] Group 4: AI Technology Development - Baidu's self-developed Kunlun chips and the launch of the third-generation Kunlun P800 chip mark significant advancements in AI computing capabilities, with widespread applications across various sectors [10][11] - The launch of the Wenxin 5.0 model, with 2.4 trillion parameters, demonstrates Baidu's commitment to leading in multi-modal AI capabilities [12] Group 5: Market Position and Future Outlook - Baidu's AI cloud service has maintained a leading market share of 24.6%, reinforcing its position as a dominant player in the AI public cloud market [14] - Analysts from major financial institutions have raised their ratings and target prices for Baidu, reflecting confidence in the company's future growth prospects in the AI sector [15][16]

Core Insights - Baidu is aggressively advancing into the AI era, showcasing transformative value through its AI products and solutions, with a focus on AI cloud and marketing services [1][5][22] Financial Performance - In Q3 2025, Baidu reported total revenue of 31.2 billion yuan, with core revenue of 24.7 billion yuan; AI cloud revenue grew by 33%, AI application revenue reached 2.6 billion yuan, and AI native marketing service revenue surged by 262% to 2.8 billion yuan [1][9] AI Product Development - Baidu has launched a series of innovative AI products, including the Kunlun chip and Wenxin 5.0, marking a transition from "intelligent emergence" to "value realization" [5][6] - The new generation of Kunlun chips (M100 and M300) is optimized for large-scale inference and multimodal model training, supporting AI technology's large-scale application [6] Market Position and Growth - Baidu's AI cloud holds a 24.6% market share in China's AI public cloud service market, leading for six consecutive years; it is the choice of 65% of central enterprises and 100% of systemically important banks [7] - The AI application segment is experiencing rapid growth, with significant revenue from flagship products like Baidu Wenku and digital employees [9][10] Digital Employee and AI Applications - Baidu's digital employees, such as the "Wu Yanzu Digital English Coach" and "Dong Dong" for China Eastern Airlines, are transforming user experiences across various industries [10][11] - The no-code platform "Miao Da" has generated over 400,000 applications, significantly lowering the development barrier for AI applications [11] Autonomous Driving and Global Expansion - Baidu's Robotaxi service, "Luo Bo Kua Pao," has achieved a global leading position, with 3.1 million service trips in Q3, a 212% year-on-year increase [13][15] - The service is expanding internationally, with plans for commercial operations in Abu Dhabi and partnerships with global ride-hailing platforms [15][16] AI Search Transformation - Baidu is transforming its search engine to integrate AI-generated content, with AI content accounting for 70% of search results by October [16][17] - The monthly active users of Baidu's AI search have grown by 18.63% quarter-on-quarter, reaching 382 million [16] Valuation and Market Reassessment - The rapid growth of Baidu's full-stack AI business and advancements in sectors like Robotaxi are leading to a reassessment of its market valuation, with multiple institutions raising target prices [19][22] - Baidu is increasingly viewed as a technology company centered on AI, spanning chips, models, applications, and autonomous driving, rather than just a traditional search engine [22]

Core Insights - The relationship between humans and AI is evolving from a simple tool-user dynamic to a deep collaborative partnership, with AI becoming a co-worker rather than just a task executor [1][3][22] - By 2030, it is predicted that over 50% of Chinese society will have adopted AI technologies [1] - AI is increasingly integrated into various aspects of daily life, enhancing productivity and personal growth [3][4] Group 1: AI Integration in Daily Life - AI is deeply embedded in daily activities such as shopping, work, and education, exemplified by the use of AI digital influencers during events like Double 11 [5][7] - The "Hui Bo Xing" digital human solution has seen over 100,000 digital hosts across multiple industries, with significant growth in sales during live broadcasts [7] - Baidu's AI search capabilities have transformed traditional search results into rich media applications, enhancing user interaction and experience [11][13] Group 2: AI as a Productivity Driver - AI is viewed as a native capability that can drive productivity rather than merely a cost, leading to a potential productivity revolution across industries [4][22] - Baidu's "Miao Da" platform allows non-programmers to create applications through natural language, showcasing AI's role in democratizing technology [13][16] - The introduction of self-evolving AI agents like "Baidu Fa Mou" aims to optimize complex decision-making processes in various sectors [17][19] Group 3: Opportunities in the Super Individual Era - Individuals are encouraged to internalize AI capabilities to thrive in the AI era, transitioning from mere interaction to deep integration [23][25] - Baidu emphasizes three key application directions for internalizing AI: replacing repetitive tasks, providing unlimited content supply through AI-generated content, and discovering new insights beyond human cognition [25][26][28] - Baidu's comprehensive AI ecosystem, from hardware to applications, positions it as a leader in enabling both individuals and enterprises to harness AI effectively [28]