北京第四範式智能技術股份有限公司 章 程 二○二五年十二月 | 第一章 | 總 | | 則 | | | | | | | 1 | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | 第二章 | | | 經營宗旨和範圍 | | | | | | | 2 | | | | | | 第三章 | 股 | | 份 | | | | | | | 3 | | | | | | 第一節 | | | 股份發行 | | | | | | | 3 | | | | | | 第二節 | | | 股份增減和回購 | | | | | | | 7 | | | | | | 第三節 | | | 股份轉讓 | | | | | | | 9 | | | | | | 第四章 | | | 股東和股東會 | | | | | | | 10 | | | | | | 第一節 | | | 東 | | | | | | | 10 | | | 股 | | | 第二節 | | | 控股股東和實際控制人 | | | | | | | 14 ...

香港交易及結算所有限公司及香港聯合交易所有限公司對本公告的內容概不 負 責，對 其 準 確 性 或 完 整 性 亦 不 發 表 任 何 聲 明，並 明 確 表 示 概 不 就 因 本 公 告 全 部或任何部份內容而產生或因倚賴該等內容而引致的任何損失承擔任何責任。 Beijing Fourth Paradigm Technology Co., Ltd. 北京第四範式智能技術股份有限公司 （於 中 華 人 民 共 和 國 註 冊 成 立 的 股 份 有 限 公 司） （股 份 代 號：6682） (1) 於二零二五年十二月十九日（星 期 五）舉行之 臨 時 股 東 會 之 投 票 表 決 結 果； (2) 獨 立 非 執 行 董 事 辭 任； (3) 委 任 獨 立 非 執 行 董 事； 及 (4) 修改公司章程 茲提述北京第四範式智能技術股份有限公司（「本公司」）日期為二零二五年十二 月三日內容有關本公司臨時股東會（「臨時股東會」）之通函（「通 函」）。除 文 義 另 有 所 指 外，本 公 告 所 用 詞 彙 與 通 函 所 界 定 者 具 有 相 同 涵 義。 日期為二零二五年十二月三日的臨時股東會通告（ ...

Investment Rating - The report assigns a "Buy" rating for the computer industry, consistent with the previous rating [2]. Core Insights - The report highlights the significant growth in the usage of the Doubao large model, with daily token usage surpassing 50 trillion, reflecting a 417-fold increase since its launch [14][47]. - The competitive landscape is shifting from business model innovation to hard technology capabilities, emphasizing the importance of foundational research and engineering in AI development [5]. - The integration of AI capabilities into various applications, such as travel planning and local services, is expected to benefit companies involved in AI chips, servers, and foundational software tools [24]. Summary by Sections 1. Doubao Large Model Token Growth - The Doubao large model's daily token usage has exceeded 50 trillion, marking a significant increase from 30 trillion in September 2025, with a monthly growth rate of 22% [14][15]. - The model's commercial viability is improving as the cost of reasoning decreases, with the latest version, Doubao 1.8, optimizing for multi-modal tasks and reducing redundant computational costs [15][20]. 2. Performance Enhancements of Doubao Large Model - The Doubao 1.8 model has shown substantial improvements in multi-modal understanding and task execution, outperforming competitors like Qwen3 in various metrics [27][32]. - New models such as Seedance 1.5 Pro and Seedream 4.5 have been introduced, enhancing capabilities in video generation and image creation, respectively [33][43]. 3. Integration of Qianwen APP with Gaode - The Qianwen APP has integrated with Gaode, enabling it to transition from understanding user intent to executing real-world services, significantly enhancing user experience in travel and local services [56][58]. - The app's public testing has resulted in rapid user adoption, with over 3 million active users within 23 days of launch [53]. 4. Investment Recommendations - Companies expected to benefit from the increase in reasoning-side computational power include AI chip and server firms like Cambricon, Inspur, and Unisoc, as well as foundational software companies like Fourth Paradigm and StarRing Technology [24].

Core Insights - Muxi Integrated Circuit (Shanghai) Co., Ltd. officially listed on the Sci-Tech Innovation Board on December 17, 2025, with a market value exceeding 300 billion yuan on its first trading day [3][4]. Investment Focus - Springhua Capital focuses on early-stage investments in cutting-edge technologies, particularly in AI and semiconductor sectors, and has made significant investments in high-performance GPU and autonomous computing systems [4][6]. - The firm has invested in various promising startups in chip design, including AI computing chips, storage chips, and integrated storage-computing solutions [6][7]. Portfolio and Collaborations - Springhua Capital has a diverse portfolio that includes investments in AI, robotics, and medical technology, with notable companies such as BAIYI ZHIHUI and XINGRUI PHARMACEUTICAL [7][11]. - The firm has established deep collaborations with top universities and research institutions, leveraging its strong alumni network to identify and support innovative entrepreneurs [10][11]. Strategic Vision - Springhua Capital aims to support companies with original innovation and disruptive technologies, believing that small, effective innovations can lead to significant global changes [7][11]. - The firm is committed to facilitating the commercialization of hard technologies and helping Chinese tech companies overcome technological barriers [11].

聚焦AI从"大模型"到"大应用"， 共探产业生态融合共生新路径。 当"炼大模型"的热潮退去 产业开始严肃讨论落地 12月12日，2025长三角人工智能产业融合共生发展大会在杭州算力小镇隆重举行。大会以"智赋万象，创新共融"为主题，汇聚了中国科学院院士何积丰， 中国工程院院士刘大响、柴天佑、戴琼海等顶尖学界权威与技术领军者。临平区委书记周徐胤、政府相关负责人、企业代表、投资机构代表等300余位嘉 宾齐聚一堂共话产业未来、共商合作机遇。 一个判断：从技术可行到商业可行 过去两年，中国AI行业最不缺的就是大模型发布会，最缺的是能算清账的商业案例。今年8月，国务院印发《关于深入实施"人工智能+"行动的意见》， 明确提出推动人工智能与经济社会各行业各领域融合。这份文件的潜台词是：技术已经够好了，现在要看谁能把它用起来。 2025年，业内普遍的共识是：中国AI产业正在经历从"炼大模型"到"炼大应用"的范式转移。技术可行性不再是瓶颈，商业可行性和应用落地才是企业竞争 的真正赛道。这场在临平举办的大会，某种程度上是这个判断的一次集体确认。 临平区前瞻布局算力产业，打造了全国首个以"算力"命名的特色小镇。本次大会的召开，既是对 ...

Core Insights - The article emphasizes the transition of China's AI industry from "large models" to "large applications," highlighting the need for practical commercial use of AI technology rather than just theoretical advancements [7][50] - The 2025 Yangtze River Delta AI Industry Integration and Coexistence Development Conference showcased the collaboration between academia, government, and industry to explore future opportunities and challenges in AI [4][50] Group 1: Conference Overview - The conference gathered over 300 participants, including top scholars, government officials, and industry representatives, to discuss the future of AI and collaboration opportunities [4][8] - The Zhejiang Turing Computing Research Institute, the conference's organizer, aims to address real challenges faced by AI companies in commercialization, such as finding application scenarios and reducing computing costs [8][48] Group 2: Keynote Speeches - Academicians presented critical insights on the future of AI, with a focus on practical challenges and the need for innovative algorithms in industrial applications [12][22] - Liu Daxiang discussed the importance of AI in developing a low-altitude economy and emergency rescue systems, emphasizing the need for a unified safety information network [14] - Chai Tianyou highlighted the necessity for new algorithms to address the dynamic nature of industrial systems, distinguishing industrial AI from general AI [16] - Dai Qionghai raised concerns about the lack of original contributions to AI research in China, stressing the importance of breakthroughs in sensors, computing power, and brain intelligence for future leadership in AI [20] Group 3: Industry Perspectives - The conference featured discussions on the practical challenges of AI commercialization, with a significant failure rate of AI solutions in enterprises due to organizational and technical gaps [26][27] - The shift from a "training cycle" to a "reasoning cycle" in AI is seen as a strategic opportunity for China to leverage its advantages in engineering and cost [29] - Real-world applications of AI in urban governance and other sectors were showcased, demonstrating the potential for AI to create significant efficiencies and innovations [30][37] Group 4: Strategic Collaborations - Six companies signed agreements to establish operations in the Computing Town, focusing on key areas of the AI industry, including chip design and smart hardware [45] - The collaborative nature of these agreements reflects the conference's theme of "integration and coexistence," emphasizing the need for a cohesive industry ecosystem to support AI deployment [45][50] Group 5: Future Directions - The Computing Town aims to serve as a model for AI industry development, with a focus on reducing computing costs and fostering innovation through a supportive ecosystem [48][50] - The conference acknowledged the ongoing challenges in transitioning AI from theoretical models to practical applications, emphasizing the need for continued dialogue and collaboration among stakeholders [50]

Investment Rating - The report maintains a "Recommendation" rating for the industry, expecting the industry index to rise more than 5% over the next 3-6 months compared to the benchmark index [39]. Core Insights - DeepSeek-V3.2 has achieved leading reasoning performance among open-source models, reaching a score of 93.1% in the AIME 2025 mathematics competition, closely following GPT-5 at 94.6% [10][14]. - The DeepSeek-V3.2-Speciale version focuses on extreme reasoning and has won gold medals in prestigious competitions such as IMO 2025 and ICPC World Finals 2025, showcasing its advanced capabilities [13][14]. - The introduction of DSA technology in DeepSeek-V3.2 significantly enhances both efficiency and performance, reducing computational complexity from O(L²) to O(Lk) [19][21]. - DeepSeek-V3.2 is the first model to integrate deep thinking with tool usage, improving its reasoning generalization ability through innovative training methods [21][25]. Summary by Sections DeepSeek-V3.2: Redefining Open-Source Model Performance - DeepSeek-V3.2 was officially released on December 1, 2025, and aims to balance reasoning ability with output length, making it suitable for everyday tasks [10]. - The model's performance in benchmark tests is comparable to that of leading proprietary models, demonstrating its competitive edge in the open-source domain [10][12]. Investment Recommendations - The report suggests focusing on specific sectors within AI, including domestic computing power and enterprise-level services, highlighting companies such as Cambricon, Alibaba, and Hikvision [27][28]. - Various application scenarios are identified, including industrial, military, medical, and legal sectors, with specific companies recommended for investment [28].

Core Viewpoint - Noumena has developed an AI-native "Growth Intelligence" system aimed at helping brands regain control over their growth in a challenging market environment [3][4][6]. Group 1: Company Overview - Noumena was founded by Pei Nasi after the successful IPO of Fourth Paradigm, focusing on creating a system that allows brands to be more proactive in their growth strategies [6]. - The company has established two main models: one that increasingly understands traffic and another that increasingly understands brands, leveraging unique training data not available to large models [7][43]. Group 2: Business Model - Noumena operates on a "heavy delivery + results-oriented" model, collaborating with clients to set goals and integrating into their operations, charging fees based on delivered results rather than software subscriptions [7][44]. - The company aims to co-create growth intelligence products with leading brands, using feedback to refine their models and potentially opening the system to more brands in the future [7][44]. Group 3: Market Position and Achievements - Noumena has partnered with several leading beauty brands, achieving significant sales growth for clients, such as increasing a national brand's annual sales from 100 million to 150 million in less than a year [7][38]. - The company has gained recognition, winning the L'Oréal BIG BANG innovation award and being selected as one of the top teams in the Baidu Wenshin Cup startup competition, validating its technological capabilities [10][12]. Group 4: Future Outlook - Pei Nasi envisions building a company with a market value exceeding 10 billion, emphasizing the importance of the next two years for rapid growth and market leadership [13][51]. - The company is focused on helping brands navigate growth challenges in the AI era, addressing the increasing anxiety among CEOs regarding growth strategies [25][28].

Investment Rating - The industry investment rating is "Overweight" (maintained) [1] Core Insights - The upcoming Volcano Engine FORCE conference is expected to showcase significant advancements in AI models and ecosystem, with a focus on the release of new members of the Doubao model family, which aims to enhance performance and reduce costs [5][6] - The conference will also highlight upgrades to Agent development tools and the expansion of the Agent ecosystem, which are crucial for integrating AI into enterprise production scenarios [7] - The report emphasizes investment opportunities in the ByteDance ecosystem and AI applications, recommending various companies that are expected to benefit from these trends [8] Summary by Sections Industry Overview - The computer industry is projected to outperform the overall market [1] - The industry has shown a significant trend with a 24% fluctuation in the past months [2] Upcoming Events - The Volcano Engine FORCE conference is scheduled for December 18-19, 2025, featuring major announcements and advancements in AI technology [5] Model Developments - The new Doubao model members are anticipated to improve capabilities while lowering costs, particularly in video generation [6] Agent Ecosystem - Upgrades to Agent development tools are expected to enhance AI development processes, making them more efficient and comprehensive [7] Investment Recommendations - The report suggests focusing on companies like Hand Information, Zhongke Chuangda, and others that are positioned to benefit from AI applications and computing power [8]

Core Insights - The report emphasizes that the computer industry is driven by model iteration, application breakthroughs, and domestic computing power leadership, with a buy rating for the sector [2][4][17]. Industry Overview - The AI infrastructure sector is transitioning from "testing and trial" to "large-scale rollout in 2026" for domestic AI chip replacements, with a narrowing gap between domestic and international leading models [8][17][26]. - The AI application sector is categorized into three types of companies: those with recognized benchmark clients but in early stages, those with long-term trends but uncertain timelines, and those leveraging AI with reachable inflection points [8][17][19]. - The EDA and R&D software sector is seeing rapid improvement in domestic product capabilities, with mergers and acquisitions expected to accelerate due to policy support [8][17][21]. - In the intelligent driving and robotics sector, the report suggests focusing on competitive chip tracks, overseas expansion, and new autonomous driving opportunities, recommending global AMR leader Geek+ [8][18][21]. - The report notes that traditional trust and innovation directions are facing funding delays, while the release of HarmonyOS PC in 2025 is seen as a catalyst for business growth and technological upgrades [8][18][21]. - The energy information sector is expected to benefit from stable capital expenditures and market reforms, although policy implementation timelines need to be monitored [8][18][21]. Key Companies - In the AI infrastructure sector, recommended companies include Cambricon, Inspur, and Unisoc, with attention to Sugon [8][19]. - For AI applications, notable companies are Jingtaikong, Rainbowsoft, and Kingdee International, with additional focus on Hancloud and Guangyun Technology [8][19][21]. - In the intelligent driving and robotics sector, key players include Geek+, Black Sesame Intelligence, and Rui Ming Technology, with recommendations to monitor Desay SV and SOTER [8][19][21]. - In the EDA and R&D software sector, companies like GigaDevice, Gexin Electronics, and BGI are highlighted, with attention to Zhongkong Technology and Haocen Software [8][19][21]. - For trust and innovation, recommended companies include Kingsoft and Softcom, with additional focus on Taiji Co. and China Software International [8][19][21]. - In the energy information sector, notable companies are Guoneng Rixin and Longxin Group, with attention to State Grid Information and South Grid Digital [8][19][22].