迅策于2025年12月30日在港挂牌上市，截至目前最新市值为163亿港元。 恒生综指下一次的定期调整生效时间为2026年3月，检讨数据周期为2025年1月1日至2025年12月31日。 据LiveReport大数据预测，中国数据基础设施及分析公司迅策(3317.HK)有望于2026年3月9日的定期检讨 中纳入港股通。 迅策是一家实时数据基础设施及分析解决方案供应商，为全行业企业提供涵盖数据基础设施及数据分析 的实时信息技术解决方案。迅策聚焦"数据计算平台 + 智能体"的全栈能力构建，成为企业级AI落地的 关键赋能者，助力各行各业实现实时智能决策。 ...

2025 企业级 AI 的边界、责任与商业模式变革。 基于2025年企业级AI产品的发布特征，按产品形态可清晰划分为四大核心类型，覆盖从基础能力底座到垂直场景落地的全链路，形成完整的AI应用生 态： 1. AI Agent 平台 很多人仍然在用 CRM、OA、ERP 的方式理解企业软件，但在 2025 年，这套分类法正在失效。 真正发生变化的，并不是 AI 更聪明了，而是软件开始直接"参与工作"。 2025 年，企业级 AI 产品市场迎来爆发式增长，SaaS 企业与 AI 原生企业共同改写行业格局。 今年的产品发布呈现鲜明的实干导向，大家都聚焦在能解决什么具体业务问题、能否嵌入企业核心流程。这一年，企业对 AI 的需求已从尝鲜试用升级为 规模化落地。AI 成为深入财务结算、销售获客、供应链协同、客服响应等核心环节的数字劳动力。无论是传统 SaaS 厂商重构产品架构，还是 AI 原生企 业补齐企业级能力，所有动作的核心都指向让 AI 真正承担具体工作、创造可量化价值。 从产品形态到商业模式，从协作逻辑到责任划分，企业级 AI 正经历全方位的结构性变革。本文基于企业软件行业全年主流产品发布实践，梳理了2025年 ...

Core Viewpoint - DingTalk has officially launched Agent OS, claiming it to be the world's first AI-driven intelligent operating system for work, marking a strategic shift from a mobile internet collaboration tool to foundational infrastructure for the AI era [1][3] Group 1: Product Overview - The core of this transformation is Agent OS, designed to run and coordinate multiple AI agents while addressing three major pain points for enterprises: data security, model execution reliability, and dependable commercial delivery [3] - DingTalk has developed a new product matrix including: - DingTalk ONE: A new interactive entry point for human-AI collaboration, integrating information across various applications and proactively notifying users of important tasks [3] - DingTalk Real: An enterprise-grade AI hardware designed to ensure data security and controllability through physical isolation and emergency mechanisms [3] - AI Search: An enterprise-level AI search engine that allows secure access to data and integrates mainstream large models for compliant Q&A [3] - Wukong: A general task processing agent capable of understanding complex instructions and autonomously planning steps to complete tasks [3][4] Group 2: Business Model Transformation - DingTalk is shifting its business model from traditional SaaS, which charges based on user accounts, to a "pay-for-results" model [5] - The company showcased several "commercially deliverable agents" that directly address enterprise pain points, such as: - An "Order Agent" that reduces order processing time from 1.4 hours to under 1 minute [5] - An "AI Travel" agent that can plan trips and compare prices in one minute, helping reduce travel costs by an average of 15% [5] - An "AI Customer Service" agent that quickly initiates conversations and identifies business opportunities [5] - The accuracy of commercial delivery for these agents is targeted to exceed 90%, linking DingTalk's value directly to enterprise outcomes [5] Group 3: Strategic Shift and Market Position - DingTalk's new strategy reflects a shift from providing tools to co-creating industry solutions with enterprises, focusing on specific sectors like manufacturing, retail, and healthcare [5] - Despite positioning itself as an "open AI ecosystem," DingTalk faces challenges in balancing its reliance on Alibaba's underlying capabilities with partnerships to maintain neutrality and openness [6] - The competitive landscape includes rivals like ByteDance's Feishu and Tencent's WeChat Work, as well as international players like Microsoft and Google, all vying for dominance in the AI-driven digital workspace [6] - The launch, named "Mulan," symbolizes a new beginning, with the CEO acknowledging that many AI products are still in early development stages, indicating that the success of this initiative will depend on market acceptance and sustainable business models [6]

01 中金研究｜宏观策略周论：汇率、股市与扩大内需 9:30-10:30 核心看点： 近期人民币走强而 A 股震荡，汇率与股市为何罕见 背离？传统分析框架为何容易误判汇率走势？居民 增收计划将如何重塑内需格局？服务消费复苏下， 哪些连锁赛道蕴含反转与高成长机遇？ 嘉宾： 刘刚 Kevin 丨中金公司研究部董事总经理、海外与 港股策略首席分析师 黄亚东丨中金公司研究部副总经理、宏观分析师 周彭丨中金公司研究部副总经理、宏观分析师 林思婕丨中金公司研究部总监、旅游酒店及餐饮首 席分析师 扫码预约 02 民银国际科技 | 科技行业 2026 年展望：企业级 AI 应用落地，开启价值新周期 11:00-12:00 核心看点： 2025年是AI大模型从"技术驱动"到"需求驱动"的转 折年，2026年将是企业级AI规模化落地的"价值兑 现年"，我们长期看好企业级AI在各行业的渗透机 遇，期待行业迈入规模化落地新阶段。 嘉宾： 15:00-16:00 核心看点： 需求、价值、供给三管齐下：1、需求驱动：大时代 塑造大机遇，2、价值驱动：潜在分红率提升，3、 供给驱动：国内反内卷加码、欧洲产能退出。 嘉宾： 李永磊丨国海证券 ...

Core Insights - The enterprise-level AI industry is entering a critical phase of value verification, focusing on whether AI technology can be transformed into practical and replicable productivity [1] - BaiRong Cloud's recognition as a leading technology breakthrough institution reflects its commitment to integrating AI into business growth [1] Group 1: AI Integration into Business Processes - BaiRong Cloud aims to transition AI from merely enhancing efficiency to becoming a core production factor integrated into organizational structures [2] - The company is advancing a technology architecture centered around agents, enabling AI to move from "assisting decision-making" to "participating in execution" [2] - The introduction of "silicon-based employees" allows for clear and stable responsibilities, significantly reducing operational costs by 60%-70% in quality inspection scenarios [2] Group 2: Long-term Value of AI in Production Systems - BaiRong Cloud believes that the explosion of enterprise-level intelligent services will hinge on the long-term value derived from integrating AI into production systems [3] - The company has developed a reusable path for enterprise-level AI implementation, allowing technology capabilities to be applied across various business lines and organizational structures [3] - BaiRong Cloud emphasizes the importance of strict business constraints and execution logic in AI ecosystem construction to ensure reliability and controllability in diverse operational environments [3]

（文章来源：第一财经） 中信证券研报表示，从OpenAI企业端AI的数据来看，2025年企业级AI处于场景探索阶段，用户数和流 量实现高增，能力平权和人员降本价值凸显，且行业整体渗透率仍有较大提升空间。展望2026年，我们 认为以强化学习技术发展为基础的Agent主线仍将持续演进，带动AI从降本到增收打开更多应用场景， 其中数据分析、代码生成、人力招聘、销售辅助、智能客服等场景需求较为清晰。建议持续关注AI在 财务、人力、销售、生产、供应链等管理软件核心模块上的商业进展。 ...

Group 1 - DeepSeek released two official models: DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, with updates available on the official website, app, and API [4] - The cancellation of the VAT exemption for contraceptive drugs and devices will directly impact production and sales companies, requiring them to reassess cost structures and pricing strategies, potentially leading to increased market prices [2] - Vanke has had 570 million yuan of equity frozen for three years, as reported by the Dongguan Intermediate People's Court [2] Group 2 - The movie "Zootopia 2" has become the highest-grossing imported film in mainland China for the year, surpassing 1.9 billion yuan in box office revenue within five days of release [6] - SF Express launched a "late delivery compensation" service, where the company will bear the costs, and couriers will not be held responsible for compensation [6] - Nestlé China confirmed that the merger of its infant nutrition business with Wyeth will not affect existing operations, with the new entity set to launch on January 1, 2026 [7] Group 3 - The Chinese express delivery industry has seen its annual business volume exceed 1.8 billion packages for the first time, reflecting a strong economic momentum [5] - The former CEO of Hema, Hou Yi, announced a new venture called "Lao Cai Rui Xuan," focusing on live-streaming sales [3] - The China CDC reported that the national flu activity has reached a moderate level, with some provinces experiencing high levels of flu activity [4]

Core Viewpoint - The market is showing strong enthusiasm for AI-related investments, particularly following the announcement of a significant change in the controlling shareholder of Zhenai Meijia, which is expected to enhance its operational capabilities through the integration of AI technology [3][10]. Group 1: Shareholder Change - Zhenai Meijia announced that its controlling shareholder will change to Tanjie Yuanqing, with a transaction cost of nearly 1.8 billion yuan [3][5]. - The acquisition process involves three steps: agreement transfer, voting rights waiver, and a partial tender offer, with Tanjie Yuanqing aiming to acquire up to 44.99% of the shares [5][6]. - Following the announcement, Zhenai Meijia's stock price surged, achieving a cumulative increase of 84.94% and a market capitalization increase of nearly 3.8 billion yuan [3][6]. Group 2: Tanjie Technology - Tanjie Technology, a key player behind Tanjie Yuanqing, is recognized as a "domestic enterprise-level AI model unicorn," with a valuation exceeding 20 billion yuan [8]. - The company specializes in creating AI agents that serve as "digital employees" for businesses, enhancing productivity and operational efficiency [11][12]. - Tanjie Technology has successfully implemented AI solutions for over 50,000 enterprises, including major companies like Alibaba and ByteDance, demonstrating significant improvements in efficiency and customer retention [12]. Group 3: AI Technology and Market Trends - Tanjie Technology's AI agents are designed to automate various business processes, significantly improving customer acquisition and operational efficiency [11][12]. - The company utilizes its proprietary platforms, "Taiqing" and "Kuanhu," to develop AI solutions that integrate industry-specific models and comprehensive business data [14][15]. - The competitive landscape for enterprise-level AI is shifting towards the integration of technology with industry applications, emphasizing the importance of creating a symbiotic relationship between human and digital employees [16].

Core Insights - The article discusses the financial performance of Fourth Paradigm in the third quarter, highlighting its transition from concept to commercial viability in the AI sector, particularly in enterprise-level applications [1][2]. Financial Performance - In the first three quarters of the year, Fourth Paradigm achieved total revenue of 4.402 billion yuan, representing a year-on-year growth of 36.8%, significantly surpassing last year's growth rate [2]. - Gross profit reached 1.621 billion yuan, with a year-on-year increase of 20.1%, and the company reported its first quarterly profit in Q3 [2][3]. Business Segments - Fourth Paradigm's business is divided into three main segments: the Prophet AI platform, SHIFT intelligent solutions, and the Shishuo AIGS service [4]. - The Prophet AI platform generated revenue of 3.692 billion yuan in the first three quarters, marking a staggering year-on-year increase of 70.1%, and accounted for 83.9% of total revenue [5]. Strategic Focus - The company is strategically focusing on the Prophet AI platform while reducing resources allocated to non-core businesses, as evidenced by a 12.5% decline in SHIFT intelligent solutions revenue and a 70.0% drop in Shishuo AIGS service revenue [5][6]. - The average revenue contribution from benchmark users in key industries increased by 71.4% to 25.49 million yuan, indicating deeper customer engagement and value extraction [6]. Market Positioning - Fourth Paradigm is positioning itself as a leader in the enterprise AI sector by addressing industry-specific needs rather than competing directly with tech giants in general AI capabilities [8][9]. - The company has developed a differentiated approach by focusing on vertical applications and building a domestic computing power ecosystem, which enhances its competitive edge [9][10]. Future Growth Potential - Fourth Paradigm has launched ModelHub XC and EngineX to address the compatibility issues between domestic computing power and models, aiming to become a key infrastructure provider in the domestic AI ecosystem [10]. - The company is also exploring new sectors such as AI in retail, energy storage, and stablecoins, which, while not yet major revenue contributors, provide significant growth potential for the next 3-5 years [11].

Market Overview - The market size of AI large model solutions in China is projected to reach 3.49 billion yuan in 2024, representing a year-on-year growth of 126.4%, while the MaaS (Model as a Service) market is expected to see explosive growth of 215.7% [2][6] - The MaaS market is anticipated to grow at a compound annual growth rate (CAGR) of 66.1% from 2024 to 2029, reaching 9 billion yuan by 2029 [6][10] Cost Challenges - High infrastructure costs are a significant barrier to the scaling of enterprise-level AI applications, with Nvidia predicting global AI infrastructure spending to reach 3-4 trillion USD by 2030, with a CAGR of 38%-46% from 2025 to 2030 [3] - The cost of training a single enterprise large model can exceed 1 million yuan, with ongoing costs during the inference phase creating a rigid expenditure burden [3][4] - Small and medium-sized enterprises (SMEs) average AI investment is only 3.2 million yuan, less than one-tenth of that of large enterprises, limiting their applications to basic scenarios like intelligent customer service [4] Technical Challenges - The complexity of enterprise-level AI applications is significantly higher than personal scenarios, particularly in terms of computing power management and model adaptation [4][5] - The coexistence of multiple chip brands, including domestic and Nvidia chips, creates a "computing island" phenomenon due to differences in instruction sets and optimization logic [5] MaaS Advantages - MaaS is seen as the optimal service model for the implementation of enterprise-level AI, providing an integrated solution that includes model repositories, inference engines, and operational tools [6][10] - The cost advantages of MaaS are evident, as it significantly reduces the overall costs of AI applications through technological optimization and model innovation [6][7] - Companies using the MaaS model report a 2-3 times higher return on AI investment compared to traditional models, with the financial sector seeing returns as high as 4 times [7] Future Trends - The MaaS market is evolving towards "intelligent integration, localization, and ecosystem development," with public cloud services becoming more prevalent among SMEs and private deployments focusing on security and customization for large enterprises [10][11] - The integration of AI agents and multimodal technologies is expected to transform MaaS services from auxiliary tools to core infrastructure for digital transformation, enabling AI to become a productivity tool accessible to all enterprises [12]