东吴证券发布研报称，看好智谱(02513)AI在本土大模型技术实力、开源生态布局以及政企本地化落地 能力方面的优势，本地化业务保持稳定增长，云端业务成为主要驱动力，并认为公司有望受益于中国大 模型行业从本地化部署向云端服务转型的长期趋势。建议关注。 东吴证券主要观点如下： 本地化部署面向政企客户，提供私有化运行和定制服务，客单价高、毛利率稳定，2025年上半年占比 84.8%、毛利率59%。云端部署通过API调用和订阅制服务企业和开发者，接入门槛低、扩展性强， 2025年上半年占比15.2%，收入占比快速提升。公司将云端业务作为长期战略重点，目标未来API收入 占比提升。开源策略有效引流，近期推出的Zcode（AI代码编辑器）和Zread（代码库分析工具）进一步 增强开发者粘性，推动付费转化。历史财务表现显示收入高增长特征。2022-2024年收入分别为0.57亿 元、1.25亿元、3.12亿元，复合增长率超过130%。2025年上半年收入1.91亿元，同比增长325%，已超 2023年全年水平。 本次港股IPO发行价116.20港元/股，全球发售3741.95万股H股，净募资约43亿港元，募资后市值约511 ...

智谱、Minimax、商汤 20260107 摘要 智谱、MiniMax 和商汤分别侧重 B 端和 C 端市场，智谱通过 MaaS 平 台赋能企业，MiniMax 则以 AI 原生应用如 Talkie 星野切入 C 端，后者 月活用户达 2000 万，贡献主要收入。 智谱一体化 MaaS 平台集成 GLM 模型及工具，支持本地和云端部署， 本地化部署收入占比约 85%，客户数和客单价逐年增长，AI 代码工具订 阅 ARR 已破亿。 MiniMax 的 AI 原生应用收入占比超 70%，海外业务发展迅速，贡献总 收入的 70%以上。开放平台通过 API 用量计费，并提供专属推理资源池 等服务。 商汤凭借计算机视觉算法优势，日新系列大模型已迭代至第六代，多模 态能力突出，并具备软硬结合的算力基础和芯片支持，专注于 B 端大模 型落地。 智谱 2022-2024 年营收复合增速达 130%，但亏损扩大，2025 上半 年净亏损 17.5 亿人民币，毛利率维持在 50%以上。MiniMax 2024 年 营收 0.3 亿美元，同比增长近 8 倍，亏损与去年基本持平，毛利率有所 提升。 Q&A 智谱、MiniMax 和 ...

大模型公司正迎来上市盛宴。1月6日最新消息，智谱认购超1000倍，MiniMax获得1209倍认购。 大模型公司商业模式已经开始清晰，一部分公司重金加注To C，另外一部分始转向To B市场。 2025年，在全球企业级LLM API市场，Anthropic凭借其Claude系列模型，以32%的生产环境使用份额击败OpenAI，跃居市 场第一。 更重要的是，Anthropic预计最早将在2027年实现现金流转正，比OpenAI的2030年目标提前三年。 AI大模型公司的盈利方式一直是市场关注的焦点。Anthropic聚焦B端客户，MaaS平台的API调用服务是其核心收入来源。 而在二级市场想要获得较高的估值，又需要向市场证明自己的赚钱能力，这就需要打通ToB的商业链路来获得盈利。 对于智谱来说，它正与时间和市场一同进行赛跑。 相比通过C端订阅和广告收入作为主要收入来源的模式——如OpenAI，Anthropic的ToB业务显得更接近稳定盈利的商业模 式。 有鉴于此，国内大模型厂商也纷纷选择跟上。 智谱作为国内同样专注ToB业务的大模型厂商，CEO在12月初也明确表示，公司计划将API业务的收入占比提升至50%。 ...

Core Viewpoint - Beijing Zhiyu Huazhang Technology Co., Ltd. (Zhiyu) is set to launch its IPO, aiming to raise approximately HKD 4.3 billion with a market valuation exceeding HKD 51.1 billion upon listing on the Hong Kong Stock Exchange [1][3] Group 1: IPO Details - Zhiyu plans to issue 37.42 million H-shares, with 1.87 million shares available for public offering in Hong Kong and 35.55 million shares for international offering [1] - The IPO is expected to conclude on January 5, 2026, with the official listing scheduled for January 8, 2026, under the stock code "2513" [1] Group 2: Investor Participation - The IPO has attracted significant interest from cornerstone investors, who are expected to subscribe to HKD 2.98 billion, accounting for nearly 70% of the total offering [3] - Notable cornerstone investors include JSC International Investment Fund SPC, JinYi Capital Multi-Strategy Fund SPC, and several prominent asset management firms [3] Group 3: Fund Utilization - Approximately 70% of the net proceeds (around HKD 2.9 billion) will be allocated to the development of general AI models, enhancing Zhiyu's competitive edge [5] - About 10% (approximately HKD 420 million) will be used to optimize the company's MaaS platform, including updates to foundational models and infrastructure [5] Group 4: Financial Performance - Zhiyu's revenue has shown significant growth, with figures of HKD 57.4 million in 2022, HKD 124.5 million in 2023, and projected HKD 312.4 million in 2024, reflecting a compound annual growth rate of 130% [7] - The company reported a revenue of HKD 190.9 million in the first half of 2025, marking a year-on-year increase of 325% [7] Group 5: Market Position - According to Frost & Sullivan, Zhiyu ranks first among independent general model developers in China and second overall, holding a market share of 6.6% based on 2024 revenue [8] - Despite revenue growth, Zhiyu has incurred significant net losses, with figures of HKD 1.43 billion in 2022, HKD 7.88 billion in 2023, and HKD 29.58 billion in 2024, alongside a projected loss of HKD 23.58 billion for the first half of 2025 [8] Group 6: R&D Investment - Zhiyu's R&D expenditures were HKD 844 million in 2022, HKD 5.29 billion in 2023, and HKD 21.95 billion in 2024, with a total of approximately HKD 4.4 billion invested in R&D as of mid-2025 [8] - The company has a high R&D personnel ratio, with 74% of its workforce dedicated to research and development [8] Group 7: Revenue Sources - The majority of Zhiyu's revenue comes from large model services, with localized deployments accounting for about 70% of total revenue, although cloud deployment revenue is on the rise [8] - The introduction of subscription products for AI programming has also contributed significantly to revenue growth, with over 150,000 paid developer users and an annual recurring revenue exceeding HKD 100 million [9]

Core Viewpoint - The company Beijing Zhiyu Huazhang Technology Co., Ltd. (referred to as "Zhiyu") is set to launch its IPO, aiming to raise approximately HKD 4.3 billion with an expected market capitalization exceeding HKD 51.1 billion upon listing on the Hong Kong Stock Exchange [1] Group 1: IPO Details - Zhiyu plans to issue 37.42 million H-shares, with 1.87 million shares available for public offering in Hong Kong and 35.55 million shares for international offering [1] - The IPO is expected to conclude on January 5, 2026, with trading commencing on January 8, 2026, under the stock code "2513" [1] Group 2: Investor Participation - The IPO has attracted significant interest from cornerstone investors, with commitments totaling HKD 2.98 billion, accounting for nearly 70% of the offering [2] - Notable cornerstone investors include JSC International Investment Fund, JinYi Capital, and various prominent asset management firms [2] Group 3: Financial Performance - Zhiyu's revenue has shown substantial growth, with figures of HKD 57.4 million in 2022, HKD 124.5 million in 2023, and projected HKD 312.4 million in 2024, reflecting a compound annual growth rate of 130% [4] - Despite revenue growth, the company has incurred significant net losses, amounting to HKD 1.43 billion in 2022, HKD 7.88 billion in 2023, and HKD 29.58 billion in 2024 [4] Group 4: Research and Development - The company has invested heavily in R&D, with expenditures of HKD 844 million in 2022, HKD 5.29 billion in 2023, and HKD 21.95 billion in 2024, totaling approximately HKD 4.4 billion in R&D investments [4] - As of June this year, 74% of the company's workforce is dedicated to research and development [4] Group 5: Market Position and Strategy - Zhiyu ranks first among independent general model developers in China and second overall, holding a market share of 6.6% based on 2024 revenue [4] - The company aims to enhance its MaaS platform revenue while maintaining its local deployment revenue base, indicating a strategic focus on scaling operations [4] Group 6: Customer Growth and Future Projections - The number of institutional clients served by Zhiyu has exceeded 12,000 in the first three quarters of this year, showing significant growth [5] - The company anticipates a revenue increase of over 60% year-on-year by the third quarter of 2025, despite expected widening net losses due to high R&D spending [5]

Core Viewpoint - The rise of intelligent agents is reshaping the dominant logic of the AI industry, transitioning from content generation to task execution, creating new competitive landscapes for model vendors and internet giants [1] Group 1: Definition and Evolution of Intelligent Agents - Intelligent agents are systems that can perceive their environment, make judgments, and take actions to achieve goals, evolving from large models initially used for text generation to more complex applications [3][5] - The emergence of intelligent agents is seen as a response to the explosion of large models like ChatGPT, prompting a reevaluation of how model companies can regain control in a rapidly changing ecosystem [3][5] Group 2: Market Dynamics and Competition - The lowering of barriers to creating intelligent agents allows a wider range of users, from casual developers to large model companies, to participate in the market, leading to a more competitive environment [6][8] - Major model vendors are transitioning from merely providing models to offering comprehensive capabilities through MaaS (Model as a Service) platforms, indicating a shift towards higher-level applications [8][12] Group 3: Industry Structure and Future Outlook - The competitive landscape is expected to consolidate, with only a few leading companies surviving in the foundational model layer, similar to the cloud computing evolution where only a handful of players dominate [11][12] - The upper layers of the market, closer to user needs, will see more diverse players due to the complexity of user demands and application scenarios, providing opportunities for differentiation [12][49] Group 4: Challenges and Opportunities for Enterprises - Enterprises are increasingly focused on the ROI of AI implementations, with a clear demand for measurable business value from AI investments [46][48] - The integration of intelligent agents into existing enterprise systems is seen as a potential solution for improving operational efficiency, although many companies still face challenges in digital transformation [32][49] Group 5: Impact on Various Industries - The software industry, particularly those focused on code models, is expected to be significantly impacted, with productivity gains from intelligent agents allowing for faster project completion [53] - Consulting and data analysis sectors may also see transformations as intelligent agents can generate comprehensive reports and analyses, although the human element in consulting remains irreplaceable [54][55]

Core Viewpoint - Beijing has introduced the "Action Plan for Supporting Information Software Enterprises to Enhance AI Application Service Capabilities (2025)", which includes multiple financial incentive policies, with a maximum reward of 30 million yuan [1]. Group 1: Financial Incentive Policies - The "soft eight measures" include six main financial policies: computing power vouchers, model "first plan", software intelligent transformation projects, data vouchers, shared open-source project rewards, and small and medium-sized enterprise service vouchers, all of which are "achievable upon meeting standards" [1]. - The computing power voucher policy supports two main areas: subsidies for computing power deployment costs for MaaS platforms and support for general intelligent agent operations, with a maximum subsidy of 30 million yuan [2]. Group 2: Software Intelligent Transformation Projects - The software intelligent transformation project policy supports two aspects: enhancing software development efficiency through AI applications and upgrading software products' intelligence levels [3]. - For software development, it encourages the use of computing power, large model deployment, and data governance to transform development methods and improve efficiency, with a maximum reward of 30 million yuan for individual enterprises [3]. Group 3: Data Voucher Policy - The data voucher policy encourages enterprises to open AI model training datasets to the public, with support based on the scale, quality, update frequency, and application effectiveness of the datasets, providing up to 500,000 yuan for individual enterprises or institutions [3]. Group 4: Implementation and Future Plans - The application details for the "soft eight measures" are currently being refined, with plans to publicly release them in the second quarter as part of the Beijing High-Precision Industry Development Project Fund Implementation Guidelines (second batch) [3].