Group 1 - Chinese artificial intelligence stocks experienced a rally, with several companies unveiling upgraded models and policymakers advocating for broader technology adoption [1][3] - Zhipu AI, trading as Knowledge Atlas Technology, saw a 30% surge in shares after launching its GLM-5 open-source large-language model, which boasts enhanced coding capabilities [1][2] - MiniMax's shares jumped 11% following the launch of its M2.5 open-source model, designed for Max coding and agentic workflows [2] Group 2 - The Shanghai STAR AI Industry Index increased by 1.7% before reducing its gains, reflecting the overall positive sentiment in the AI sector [2] - DeepSeek upgraded its flagship AI model, adding support for a larger context window and more current knowledge, indicating ongoing competition among Chinese developers [3] - Ant Group released its open-source AI model, Ming-Flash-Omni 2.0, which can generate speech, music, sound effects, and visuals, showcasing advancements in multimodal AI capabilities [4] Group 3 - Chinese Premier Li Qiang emphasized the need for a comprehensive push to implement AI across diverse scenarios to unlock its potential [4]

Investment Rating - The report assigns a "Buy" rating to the company [3] Core Insights - The company is a leading provider of AI large model services in China, focusing on the development of general large models (AGI) and has established a comprehensive model product matrix, including multi-modal, agent, and coding models [15][20] - Revenue is expected to grow rapidly, with projections of RMB 7.4 billion, RMB 15.3 billion, and RMB 28.8 billion for 2025, 2026, and 2027 respectively, reflecting growth rates of 138.1%, 106.2%, and 87.9% [18] - The company has a strong emphasis on R&D, with expenses increasing significantly, indicating a commitment to enhancing model competitiveness [17][50] Summary by Sections Company Overview - The company was founded in 2019 and has positioned itself as a leading AI company in China, focusing on the development of general large models [20] - It has supported over 8,000 institutional clients and covered approximately 80 million devices as of mid-2025 [20] Revenue Growth - Revenue has expanded from RMB 60 million in 2022 to RMB 310 million in 2024, with a compound annual growth rate exceeding 130% [15][37] - The majority of revenue comes from localized deployment, which accounts for over 80% of total revenue, driven by demand for data security and compliance [16][37] R&D Investment - R&D expenses surged from RMB 80 million in 2022 to RMB 1.6 billion in the first half of 2025, representing 835% of revenue [17][50] - The company is transitioning its R&D cost structure from labor-intensive to compute-intensive, reflecting a shift in focus towards computational resources [54] Profitability Forecast - The company is projected to have negative earnings per share (EPS) of -7.79, -8.14, and -7.26 for 2025, 2026, and 2027 respectively [18] - The report estimates a reasonable total value of RMB 107.36 billion, corresponding to a fair value of HKD 273.22 per share [18] Market Position - The company ranks first among independent general large model developers in China based on revenue for 2024, with a market share of approximately 6.6% [20] - The competitive landscape in the AI large model market is fragmented, and the company aims to enhance its market share through significant R&D investments [50]

Core Viewpoint - The company, Zhipu, is poised for significant advancements in artificial general intelligence (AGI) with the upcoming release of its new model, GLM-5, and a strategic focus on innovative model architectures and learning paradigms [2][3][12]. Group 1: Company Development and Achievements - Zhipu has successfully launched multiple models, including GLM-4.1, GLM-4.5, GLM-4.6, and GLM-4.7, achieving significant improvements in model capabilities and user engagement [11][12]. - The company reported a remarkable increase in annual recurring revenue (ARR) from 20 million to 500 million (25 times) within 10 months, with overseas revenue exceeding 200 million [11]. - Zhipu's model GLM has been recognized as a national model in Malaysia, indicating its growing influence and acceptance in international markets [11]. Group 2: Strategic Focus and Future Directions - The company aims to focus on four key areas by 2026: the launch of GLM-5, the design of new model architectures, enhanced reinforcement learning (RL) capabilities, and the exploration of online and continual learning paradigms [3][13]. - Zhipu has established a new department, X-Lab, dedicated to fostering innovation and collaboration among young talents, focusing on groundbreaking explorations in AI [4][14]. - The company emphasizes the importance of long-term goals over short-term gains, aiming to contribute to global technological progress and user-centric AI solutions [7][10]. Group 3: Industry Context and Vision - The AI industry is expected to experience a significant transformation, with Zhipu positioning itself as a leader in this evolution by developing models that can replace various job functions and tasks [3][12]. - The company envisions a future where AI fundamentally alters computing paradigms, potentially leading to a new generation of AI-centric computing systems [9][10]. - Zhipu's commitment to AGI reflects a broader industry trend towards open, transparent, and accessible AI technologies that benefit all users [10][12].

Core Insights - The article discusses the emergence of AI2C (AI to Consumer) application companies in China, highlighting the potential for a billion-dollar market opportunity that is ready to be unlocked [3][4]. Group 1: Event Overview - The event featured deep discussions with five AI entrepreneurs and experts, focusing on software applications from models to production, and strategies for going global [3][5]. - A total of 40 seats were available for this closed-door event, emphasizing exclusivity and targeted networking [1][12]. Group 2: Entrepreneur Presentations - Four early-stage entrepreneurs presented their business models in five-minute pitches, aiming to identify the first true demand in AI2C and strategies for international expansion [5][6]. - Notable entrepreneurs included: - The founder of Xiaoying Technology, which has developed popular video creation software with over 100 million global users in less than two years [12]. - The CEO of YouMind, an AI learning and creation tool that aims to create a comprehensive creative ecosystem [14]. - The head of Z.ai, which has seen a tenfold increase in paid users within two months after launching new models [15]. - The head of Amazon Web Services in Greater China, which offers over 240 services across 120 availability zones globally [16]. Group 3: Roundtable Discussions - The event included interactive roundtable discussions where participants could engage directly with mentors on specific topics, fostering deeper insights into the AI2C landscape [10][11]. - The format allowed for personalized discussions, enhancing the learning experience for attendees [11][12]. Group 4: Collaboration and Support - The event was supported by various organizations, including Amazon Web Services and Linkloud, indicating strong industry backing for AI entrepreneurship [22][23]. - Collaboration with academic institutions like Tsinghua University highlights the integration of research and practical applications in the AI sector [23].

Core Viewpoint - The article discusses the recent layoffs at Zhipu, a leading AI company, highlighting the urgency and lack of empathy in the process, which reflects the company's struggle between profitability and the pressures of going public [4][11][20]. Group 1: Layoff Details - Zhipu has implemented rapid layoffs, with employees given only two days' notice before their last working day, which has caused significant frustration among staff [5][8]. - The layoffs affected key departments, including the R&D center, which is crucial for the company's revenue generation, particularly from state-owned enterprises [7][9]. - Employees believe the layoffs are part of a strategy to present a more favorable financial report ahead of the company's IPO, aiming for higher efficiency [9][19]. Group 2: Organizational Restructuring - The layoffs are seen as a compromise following a loss of momentum in commercialization and imbalances in investor power [11][12]. - Following a major strategic shift in April, Zhipu restructured its business lines to reduce overlap and improve efficiency, but this did not initially lead to layoffs [12][14]. - The restructuring was influenced by the success of DeepSeek, which has changed the competitive landscape for AI companies, forcing Zhipu to adapt its business model [14][15]. Group 3: IPO Preparation - Zhipu is the first among the "AI Six Dragons" to pursue an IPO, with plans to submit its application by October [17][18]. - The company has faced criticism for its organizational structure, which is perceived as top-heavy, necessitating adjustments before the IPO [18][19]. - The adjustments are aimed at correcting the valuation logic in the capital market, moving away from low-margin, long-cycle customized projects to more profitable models [20][21]. Group 4: Market Position and Challenges - Zhipu's shift towards C-end products, particularly focusing on international markets, is part of its strategy to enhance profitability and market appeal [20][21]. - The company faces significant competition from established players and emerging startups, making it challenging to carve out a distinct market position [22].

Core Viewpoint - The Z DemoDay showcased 24 early-stage AI projects, marking a shift from "tool revolution" to "intelligent agent revolution" in AI entrepreneurship [6][10]. Group 1: Event Overview - The event was hosted by Xinglian Capital (Z Fund) and ZhiPu Z Plan, featuring 24 AI projects across various sectors including finance, law, embodied intelligence, education, animation, and productivity tools [2][6]. - The event attracted 450 investors and emphasized the importance of diverse innovation elements to increase the probability of breakthroughs in AI entrepreneurship [6][10]. Group 2: Project Highlights - The showcased projects included: - **AmiO Robot**: Focused on intelligent manufacturing solutions [36]. - **Xinyan Weimo**: A platform for dynamic monitoring using AI and advanced microscopy [38]. - **LATIOS.AI**: An agent providing overseas investment intelligence [40]. - **ClipCap**: A video creation agent that streamlines the video production process [43]. - **ComindX**: A personalized cognitive engine for content creation [45]. - **Edison**: An AI video growth agent targeting content creators [48]. - **Fortunetell AI**: Offers decision support based on traditional wisdom and AI [50]. - **iOffer**: An AI-driven study abroad application assistant [53]. - **Jing Tong Technology**: A platform for AI knowledge management [55]. - **Lovpen**: An AI agent for content creation tailored for KOLs [58]. - **PowerLaw**: A legal tech company providing AI-driven contract solutions [61]. - **Mob.AI**: Focused on creating engaging virtual experiences [63]. - **Perle**: An entertainment tech company leveraging AI for content creation [65]. - **Singularity**: An AI hardware company for children's education [68]. - **Qujing Technology**: A pioneer in model inference acceleration [71]. - **Tian Da Zhi Tu**: A knowledge base for traditional Chinese medicine [74]. - **Xing Po Yun Miao**: An AI animation production platform [76]. - **BeingBeyond**: A company focused on embodied intelligence [79]. - **6E**: A GenAI solution provider in Southeast Asia [83]. - **LemonAI**: A self-evolving AI agent [85]. - **Mingzhi AI**: An AI-native company focused on enterprise applications [87]. - **Prismer**: A research platform for elite researchers [97]. - **Shen Yuan Zhi Yao**: A biotech company focused on AI-driven drug development [98]. Group 3: Innovation and Collaboration - The event highlighted the importance of creativity in AI, showcasing how AI can serve as a catalyst for human imagination and innovation [27][30]. - The collaboration between Z Plan and Xinglian Capital aims to build an "AI creativity symbiosis network," providing entrepreneurs with technological empowerment and capital acceleration [30].

作者｜ Stone Jin 微信公众号｜ipozaozhidao 第一家正式启动IPO流程的"大模型创业公司"。 本文为IPO早知道原创 据 IPO早知道消息， 北京智谱华章科技股份有限公司 （以下简称 " 智谱 "）于2025年3月31日同中 金公司签署辅导协议，正式启动 A 股 IPO进程。 这意味着，智谱成为 "大模型创业公司"中第一家正式启动上市流程的企业 。 成立于 2019年的智谱 致力于打造新一代认知智能大模型 。早在 2020年 年 底 ，智谱就研 发 了 GLM预训练架构， 并于 2021年训练完成百亿参数模型GLM-10B，同年利用MoE架构成功训练出收 敛的万亿稀疏模型，2022年研发了中英双语千亿级超大规模预训练模型GLM-130B并开源。2023 年，智谱推出千亿基座对话模型ChatGLM并两次升级，开源版本的ChatGLM-6B让大模型开发者的 本地微调和部署成为可能 。 2024年1月，智谱推出新一代基座大模型GLM-4，整体性能相比上一代大幅提升；6月开源GLM-4- 9B及视觉模型GLM-4V-9B，多模态能力媲美GPT-4V；7 月推出视频生成模型CogVideoX，推理速 ...