机构：开源证券 研究员：初敏/荀月 云、芯片、Robotaxi 驱动价值重估，维持"买入"评级短期搜索业务承压，但随着文心大模型对C 端应用 的重构、智能云在MaaS 市场的份额提升、昆仑芯在国产算力替代中的战略地位，以及萝卜快跑 （Apollo Go）在武汉等地商业模式跑通，2026 年有望成为百度AI 商业化变现的拐点。我们下调公司 2025-2027 年non-GAAP 净利润预期至185/200/224 亿元（前值： 196/220/248 亿元），对应同比增速-31.6%/+8.6%/+12.0%，对应EPS 6.4/7.0/7.8 元，当前股价对应 PE19.6/18.1/16.2 倍。考虑到昆仑芯拆分上市有望驱动估值提升、AI 云收入贡献持续扩大、智能驾驶积 极海外扩张，维持"买入"评级。 AI：从底层芯片到顶层应用高效协同 智能云端到端优化成为差异化优势，昆仑芯2026 年有望进入放量期。2025Q3 百度智能云收入62 亿元， YOY+21%，其中AI 基础设施、平台服务及公有云收入42 亿，YOY+33%。百度在AI 四层架构（芯片- 框架-模型-应用）均拥有自主可控核心技术。芯片层，昆仑 ...

Core Insights - Baidu Smart Cloud launched the OpenClaw one-click deployment service, enabling developers to quickly deploy AI digital employees with minimal cost [1][3] - OpenClaw's core value lies in its "embodied operation" capability, allowing it to perform tasks like a human, thus reducing labor costs and improving office efficiency for enterprises [1][2] Deployment and Integration - Developers can deploy OpenClaw in just a few minutes using Baidu Smart Cloud's lightweight application server, significantly lowering the technical barrier for users without deep technical backgrounds [1] - OpenClaw integrates seamlessly with Baidu's Qianfan platform, allowing for flexible switching between various mainstream AI models based on task requirements [2] Free Trial Promotion - To promote the adoption of AI agents, Baidu Smart Cloud is offering a limited-time free trial starting January 31, allowing users to experience OpenClaw at no cost for the first month [3] - The promotion includes a daily limit of 500 units, encouraging developers and enterprises to explore AI applications and enhancing the overall accessibility of AI technology [3]

Core Insights - Baidu Intelligent Cloud has made significant progress in its ecosystem construction with a "partner-first" strategy, achieving over 100% year-on-year growth in core partner revenue and a threefold increase in revenue from channel partners [1] - The company aims to provide a comprehensive AI toolkit to partners, facilitating easy integration into their products and solutions, thereby capturing the benefits of intelligent transformation [1] Group 1: Ecosystem Development - The company is building a "big ecosystem" that integrates four types of partner capabilities: product, opportunity, delivery, and channel, to create measurable value [2] - Product partners are converting industry knowledge into standardized solutions, exemplified by a collaboration with Xiamen Zhongshu Technology to create an intelligent procurement platform [2] Group 2: Revenue Growth - Opportunity partners have achieved significant growth, with total amounts from collaborations with Baidu Intelligent Cloud reaching several billion [3] - Channel partners have experienced rapid growth, with revenue from million-level partners increasing by 331% year-on-year, and a notable example is Henan Tengyou Technology, which saw a fourfold increase in overall performance after expanding its role [3] Group 3: Industry Innovation - Baidu Intelligent Cloud is collaborating with partners to define new industry standards through deep capability integration and joint innovation [4] - In the smart transportation sector, a partnership with Hubei Taiyue Satellite has led to a significant reduction in average vehicle delays by 8% in key urban areas, enhancing traffic efficiency [4] - In the smart elderly care sector, a project with Rui Lian Kang Da has established a scalable model for home-based elderly care, improving service response efficiency and user engagement [4] Group 4: Future Directions - The company plans to increase ecosystem investments in 2026, focusing on providing a richer product offering, building more open collaboration policies, and creating more integrated platform capabilities [5] - The overarching goal is to share greater growth dividends with partners by constructing a "value community" to address the opportunities presented by AI industrialization [6]

Core Insights - Baidu has launched the official version of its native multimodal large model, Wenxin 5.0, which features 2.4 trillion parameters and supports various forms of input and output, including text, images, audio, and video [1] Group 1: Model Technology - Wenxin 5.0 utilizes a unified autoregressive architecture for native multimodal modeling, allowing for joint training of multiple data types within a single framework, enhancing the integration and optimization of multimodal features [1] - The model employs a large-scale mixture of experts (MoE) structure with ultra-sparse activation parameters, activating less than 3% of parameters, which improves inference efficiency while maintaining strong model capabilities [1] - The model's training incorporates long-term task trajectory data and end-to-end reinforcement learning based on thinking and action chains, significantly enhancing its agent and tool invocation capabilities [1] Group 2: Application Development - Baidu has developed matrix and specialized models based on the Wenxin foundational model, with matrix models aimed at product-level applications and general scenarios, while specialized models target industry applications and vertical scenarios [4] - Three technological breakthroughs were highlighted: an end-to-end synthesis model based on voice tokens, a technology that surpasses human live streaming in 5 minutes, and real-time interactive digital human technology [4][5] - The real-time interactive digital human technology employs a three-state token linkage architecture for streaming control of text, voice, and video, allowing for automatic action transitions and low-latency, high-expressiveness video output [5] Group 3: Platform and Agent Development - Baidu's Qianfan platform has developed over 1.3 million agents, with daily tool invocation exceeding ten million times, showcasing the platform's extensive application and integration capabilities [6] - The platform offers comprehensive support for enterprise operations across multiple scenarios, significantly lowering the innovation threshold for enterprise agents [5][6]

Core Viewpoint - The report from Jefferies highlights Baidu's launch of ERNIE 5.0 and its comprehensive Qianfan intelligent agent platform, addressing developers' pain points in model usage, tool application, intelligent agent development, data services, and operational aspects [1] Group 1: ERNIE 5.0 Launch - Baidu officially released ERNIE 5.0, showcasing strong performance in multimodal understanding, creative writing, and instruction adherence, utilizing a mixed expert architecture with 2.4 trillion parameters [1] - The pre-training cycle speed has improved by 240%, integrating diffusion model-based generation technology for high-quality output [1] Group 2: Qianfan Intelligent Agent Platform - The Qianfan platform offers a comprehensive intelligent agent solution, focusing on overcoming challenges related to performance, capability, and security [1] - It boasts 1.3 million intelligent agents with daily usage exceeding tens of millions [1] Group 3: Industry Positioning - Jefferies believes Baidu demonstrates a leading position in multimodal AI technology and intelligent agent infrastructure through ERNIE 5.0 and the Qianfan platform [1] - The application of models across various industries promotes deep integration of AI technology with industry, providing robust support and innovative capabilities for developers and enterprises [1]

Core Insights - The AI landscape in 2025 is characterized by intense competition and significant advertising presence, particularly in major cities like Beijing, Shanghai, Guangzhou, and Shenzhen, where cloud service providers dominate advertising spaces [1][2] - The focus has shifted from merely showcasing AI capabilities to demonstrating practical applications that solve real business problems, indicating a transition from a storytelling phase to a results-driven phase [3][4] Group 1: Market Dynamics - Baidu Intelligent Cloud has secured 95 bidding projects with a revenue of 710 million yuan, significantly outperforming competitors like Volcano Engine and Huawei Cloud, which earned 475 million yuan and 446 million yuan respectively [2][3] - The disparity in revenue highlights Baidu's strategic focus on practical applications of AI rather than just theoretical models, positioning it as a leader in the market [3][4] Group 2: Business Transformation - Companies are misinterpreting AI transformation as merely implementing chatbots, while true transformation requires integrating AI into core business processes to enhance productivity [7][8] - Successful examples include China Railway and Taikang Insurance, where AI has drastically reduced operational time and improved training efficiency, showcasing the tangible benefits of AI integration [10][13] Group 3: Infrastructure and Cost Efficiency - Baidu's AI infrastructure is designed to optimize performance and reduce costs, with a focus on minimizing data transmission losses and maximizing training efficiency [16][18] - The dual-layer infrastructure, comprising AI Infra and Agent Infra, allows for seamless integration and deployment of AI solutions, making it easier for businesses to adopt and benefit from AI technologies [19][22] Group 4: Industry Adoption - A significant portion of China's central enterprises and major banks are utilizing Baidu Intelligent Cloud, indicating a strong trust in its capabilities to handle critical operations [24][25] - The emphasis on reliability and stability in AI applications is crucial for industries such as finance and energy, where operational continuity is paramount [25][26] Group 5: Long-term Vision - Baidu's strategic investments in AI infrastructure since 2011 reflect a long-term vision that prioritizes foundational capabilities over short-term gains, positioning it well for future challenges [30][31] - The focus on building a robust AI ecosystem that integrates deeply with industrial applications is seen as essential for sustainable growth and competitiveness in the evolving AI landscape [31][32]

Investment Rating - The report maintains a "Buy" rating for Baidu Group-SW (9888.HK) [1] Core Views - Baidu's Q3 2025 revenue was approximately 31.17 billion yuan, a year-on-year decline of 7% and a quarter-on-quarter decline of 5%. The non-GAAP net profit was 3.77 billion yuan, down 36% year-on-year and 21% quarter-on-quarter, with a gross margin of approximately 41.25% [4][5] - The core business revenue for Baidu was 24.66 billion yuan, also down 7% year-on-year and 6% quarter-on-quarter. The non-GAAP operating profit was 2.23 billion yuan, a significant drop of 67% year-on-year and 49% quarter-on-quarter [4][5] Summary by Sections Recent Performance - Baidu's total revenue for Q3 2025 was approximately 31.17 billion yuan, with core revenue of about 24.66 billion yuan, both reflecting a 7% year-on-year decline. iQIYI's revenue was 6.7 billion yuan, down 8% year-on-year [4][5] - Online marketing revenue decreased by 18% year-on-year, while non-online marketing revenue grew by 21% [5] Online Marketing - The online marketing segment saw revenue of approximately 15.3 billion yuan, down 18% year-on-year. Baidu is advancing AI transformation in its search services, with about 70% of mobile search results pages containing AI-generated content. AI-native marketing services have become a significant growth driver, achieving revenue of 2.8 billion yuan, a year-on-year increase of 262% [5][6] Cloud Services - Baidu is accelerating the development of enterprise AI-native applications, with AI application revenue reaching 2.6 billion yuan, a 6% year-on-year increase. The intelligent cloud infrastructure revenue was 4.2 billion yuan, up 33% year-on-year, and AI high-performance computing facility subscription revenue grew by 128% [6][7] Autonomous Driving - Baidu's autonomous driving business is expanding globally, with over 250,000 weekly orders for fully autonomous operations as of October 2025. The service has been launched in 22 cities, with significant partnerships in Switzerland and Abu Dhabi [6][7] Financial Forecast - The report adjusts Baidu's revenue forecasts for 2025-2027 to 130.2 billion, 136.8 billion, and 145.4 billion yuan, respectively. Non-GAAP net profits are projected at 19.2 billion, 20.1 billion, and 22.7 billion yuan, with corresponding non-GAAP EPS of 6.99, 7.29, and 8.26 yuan. The non-GAAP P/E ratios are expected to be 14.9, 14.3, and 12.6x [7][9]

Core Insights - Baidu reported a strong Q3 2025 performance with total revenue of 31.2 billion yuan and core revenue of 24.7 billion yuan, driven by over 50% growth in AI business revenue, reaching 10 billion yuan [2] - The company emphasizes the transformative value of AI across its business, with a commitment to continue AI innovation to create significant value for users, enterprises, and society [2] AI Cloud Business - Baidu's AI cloud revenue grew by 33% year-on-year in Q3, with AI high-performance computing subscription revenue increasing by 128% [3] - The company detailed its AI business revenue structure for the first time, reporting AI application revenue of 2.6 billion yuan and a 262% increase in AI-native online marketing service revenue to 2.8 billion yuan [3] - Baidu maintains a leading position in the AI public cloud market with a 24.6% market share, according to IDC, marking six consecutive years as the market leader [3] Platform Strategy - The platform strategy is crucial for Baidu's smart cloud to expand its ecosystem, with the Qianfan platform now serving over 460,000 enterprise users and over 1.3 million agents developed [4] - Continuous iteration of the technology base supports cloud business growth, with the release of the new Kunlun chip and Tianchi supernode products [4] - Baidu has established a comprehensive self-research barrier in the AI sector, creating a sustainable business model through a three-pronged strategy of infrastructure, platform ecosystem, and industry scenarios [4] AI Application Growth - Baidu's AI restructuring strategy shows significant progress across multiple product lines, with AI search monthly active users reaching 382 million, a 18.63% quarter-on-quarter increase [5] - In content creation and knowledge management, Baidu's AI transformation in Baidu Wenku and Baidu Wangpan has led to over 80 million active users for Baidu Wangpan [5] - The digital human technology from Huibo Star demonstrated strong commercial capabilities during the "Double 11" shopping festival, with GMV increasing by 91% [5] Autonomous Driving - Baidu's autonomous driving service, "Luobo Kuaipao," achieved 3.1 million global ride services in Q3, a 212% year-on-year increase, with over 25,000 fully autonomous rides per week as of October [6] - The total mileage for autonomous driving exceeded 240 million kilometers, with a significant expansion into 22 cities globally [6] - The release of the Wenxin 5.0 model, featuring 2.4 trillion parameters, enhances the company's capabilities in multimodal understanding and generation [6]

Core Insights - The article discusses the evolution of AI from model competition to a comprehensive infrastructure competition, emphasizing the importance of computing power as the "new oil" [2][3]. Group 1: AI Infrastructure - The AI infrastructure consists of a full chain including chips, computing power, models, platforms, applications, and commercialization ecosystems [3]. - Self-developed chips are crucial for optimizing computing power, reducing costs, and ensuring better integration with business needs [4][5]. - Cloud service providers play a key role in AI infrastructure, with AI cloud being an advanced version that integrates AI chips, models, algorithms, and application ecosystems [7]. Group 2: Market Dynamics - The global cloud computing market is growing rapidly, with China's annual growth rate reaching 35%, and AI-related revenues are increasing at an even faster pace [8]. - In the AI public cloud segment, Baidu Smart Cloud holds a 24.6% market share and has become the first domestic provider to announce profitability in AI cloud services [8][10]. Group 3: AI Chip Development - Baidu has developed the Kunlun chip, optimized for large-scale inference scenarios, with new models like M100 and M300 being introduced [11]. - Baidu has also made advancements in multi-card clusters for the MOE architecture, with products like "Tianchi 256" and "Tianchi 512" capable of training trillion-parameter models [12]. Group 4: Comprehensive AI Ecosystem - Baidu has established a complete domestic stack system for AI, integrating Kunlun chips and the Baijiao AI computing platform for efficient task management [16]. - The company serves over 460,000 enterprise clients and has developed more than 1.3 million platform agents to assist various industries [18]. Group 5: User-Friendly AI Development - The newly launched "Miaoda 2.0" platform allows non-technical users to create applications without programming knowledge, enhancing accessibility [19]. - Baidu's Qianfan platform integrates the latest Wenxin large model and over 150 mainstream models, providing a rich development ecosystem [20]. Group 6: Competitive Landscape - The article compares Baidu's comprehensive layout to Google's, highlighting the importance of having full-stack capabilities in the AI cloud market [25]. - Companies with full-stack capabilities will be able to provide powerful, stable, low-cost, and controllable AI computing power, becoming the digital foundation of the intelligent era [25].

Core Viewpoint - The article emphasizes the importance of Baidu's long-term strategy in AI development, focusing on building foundational capabilities rather than seeking immediate gains in the fast-paced internet culture [5][12][29]. Group 1: AI Strategy and Development - Baidu's AI strategy is characterized by a commitment to "long-termism," aiming to integrate AI as a fundamental capability across various industries rather than just a tool for quick monetization [5][12]. - The company has made significant investments in foundational technologies, including self-developed chips and AI frameworks, to support its AI applications [14][20]. - Baidu's approach contrasts with the prevailing "speed wins" mentality of the internet era, suggesting that AI competition may not follow the same rules as previous tech races [9][12]. Group 2: Technological Advancements - Baidu has introduced several AI applications, such as the "Luo Bo Kua Pao" autonomous vehicle service, which has completed over 17 million rides, demonstrating the viability of autonomous driving technology [4][18]. - The company has developed the Kunlun chip, capable of supporting multiple large models simultaneously, showcasing its commitment to self-sufficiency in computing power [14][22]. - Baidu's deep learning platform, PaddlePaddle, represents a significant step towards creating a competitive domestic AI framework, challenging the dominance of international platforms [14][22]. Group 3: Market Position and Future Outlook - Baidu's focus on "internalizing AI capabilities" aims to create a sustainable competitive advantage by embedding AI deeply within its business processes [13][24]. - The company is positioning itself as a "military supplier" of AI capabilities, offering modular solutions to businesses, which could lead to widespread adoption of AI across various sectors [24][26]. - Baidu's long-term investments and strategic choices are expected to yield significant returns as the AI landscape matures, moving from a focus on individual breakthroughs to a comprehensive ecosystem of AI applications [28][29].