IPO 不是对胜者的奖赏，而是下一轮竞赛的鼓点。 文 丨 程曼祺 编辑 丨 宋玮 接连两天，大模型创业公司智谱和 MiniMax 港股 IPO。对比移动互联网的几次上市盛宴，大模型领域的 IPO 并不发生在大战告一段落之后。它不是对胜者的奖赏，而是下一轮竞赛的鼓点。 在智谱和 MiniMax 前后脚登陆二级市场后，他们将开启更大规模的定增。这是一个商业化仍不确定、 持续的研发投入却十分确定的领域。 IPO 的实质意义是更高效地获得更多资源。 MiniMax 上市前夕，我们采访了 MiniMax 团队和他们的多位投资人，共同还原过去 3 年多里，市场对 大模型创业机会的多种视角，以及这家公司的特质。 上市前的 7 轮融资中，30 家机构共投资 MiniMax 15 亿美元。阿里投了最多的钱；高瓴是第一轮领投 方，按份额计算仅次于阿里，是第二大外部股东；明势参与了最多轮次。 在今天（1 月 9 日）早上前往港交所敲钟前，MiniMax 创始人闫俊杰对《晚点 LatePost》分享了他此刻 的想法： 希望我们后续能有机会对整个行业智能水平的提升做出更大的贡献。我们初步探索了一条纯草根 AI 创业的路径，尽管后面还是 ...

1月8日，"全球大模型第一股"智谱在港交所主板挂牌上市，发行价为每股116.20港元，以120港元/股的价格开盘后，最高冲至130港元/ 股， 涨幅超过10%，市值突破570亿港元。 在业内看来，智谱成功叩开港交所大门，是中国人工智能产业尤其大模型领域从技术探索迈向规模化商业应用的重要里程碑。智谱的上 市，不仅为全球投资者提供了分享中国AI基础层技术红利的机会，更有利于重塑中国科技股在资本市场的估值逻辑。 何以被OpenAI列为竞争对手？ "全球大模型第一股"诞生！ 全球AI巨头Open AI曾在一份名为《Chinese Progress at the Front》（中国在前沿领域进展）的官方报告中提到，一家名为"智谱"的公司 取得了显著进展。围绕主权AI竞争，OpenAI认为智谱已经成为自己的重要对手之一。 Open AI点名的这家公司，即为中国最大的独立大模型厂商——北京智谱华章科技股份有限公司（以下简称"智谱"或"智谱华章"），是一家 以AGI（通用人工智能）基座模型为核心业务的独角兽。 智谱是中国最早研发大模型的企业，被业界称为中国大模型的"黄埔军校"。 智谱成立于2019年，由清华大学技术成果转化而 ...

12月21日，港交所官网显示，中国AI大模型龙头企业MiniMax（稀宇科技）发布聆讯后资料集（PHIP）版本的招股书资料，正式冲刺港股"大模型第一 股"，有望刷新纪录，成为从成立到IPO历时最短的AI公司。 | 壽纂]的[編纂]數目 : | | [編纂](視乎[編纂] | | --- | --- | --- | | | | 行使與否而定) | | 「編纂]數目 | . . | [編纂](可予重新分配) | | [編纂]數目 | .. | [編纂](可予重新分配及 | | | | 視乎[編纂]行使與否而定) | | 最高[編纂] | … | 每股[編纂]港元,另加1% | | | | 經紀佣金、0.0027%SFC交易徵費、 | | | | 0.00565%聯交所交易費及0.00015% | | | | 會財局交易徵費(須於[編纂]時以港元 | | | | 繳足,多繳股款可予退還) | 招股书显示，MiniMax成立于2022年初，是一家全球化的通用人工智能（AGI）公司，也是少数自创立起就专注于全模态模型研发的大模型公司之一， 以及当前国际化收入最高的中国大模型公司。 这是一支年轻的团队，由385人组成，平 ...

来源：格隆汇APP 格隆汇12月21日｜中国"AI六小虎"、人工智能初创公司稀宇科技(MiniMax)在港交所上载聆讯后资料 集，拟根据《上市规则》第18C章在香港主板上市，有望成为从成立到IPO历时最短的AI公司。 该公司的大模型组合包括大语言模型、视频生成模型以及语音和音乐生成模型。MiniMax M系列由 MiniMax M1和MiniMax M2组成，是其大语言模型旗舰产品线。MiniMax M1今年6月推出，是一款开源 的大规模混合注意力推理模型。 Hailuo-02系列模型能够从多种形式信息输入中生成高质量视频内容。Speech-02模型旨在从文本输入中 生成自然、高质量的语音。MiniMax是公司的智能agent应用，可通过自然语言指令自主执行类型广泛的 任务。该公司收入主要有两大来源：(i)AI原生产品及(ii)开放平台及其他基于AI的企业服务。 稀宇科技是通用人工智能科技公司，于2022年初成立，致力推动人工智能科技前沿发展，实现通用人工 智能(AGI)。该公司投资者包括阿里巴巴及腾讯。 ...

文本模型上，2025年10月，公司发布并开源新一代文本大模型MiniMax M2，发布即在Artificial Analysis 位列全球前五、开源第一，也是中国开源大模型首次在该榜单中跻身全球前五。M2在全球模型聚合平 台 OpenRouter上迅速爬升到国内模型 token 用量第一，编程场景排名全球token用量第三。 在持续高强度研发投入、快速迭代全模态模型的同时，MiniMax的商业化效率与组织效率都尤为高效。 根据招股书数据显示，经调整净亏损在2025年与上年同期相比近乎持平，实现了在高速增长下的亏损有 效收窄。这得益于多元化的收入模型与高效的费用投入——2025年前九个月，在收入同比增长超170% 的同时，研发开支同比增幅为30%，销售及营销开支更是同比下降26%，印证了其产品依靠模型智能与 用户口碑驱动的增长逻辑，而非依赖巨额流量投入。 更让人震惊的是，MiniMax自成立到25年9月累计花费5亿美金（约35亿RMB），对比OpenAI的400亿至 550亿美元累计花销，MiniMax仅仅用了不到1%的钱做了全模态全球领先的公司。 截至2025年9月30日，MiniMax已有超过200个国家及 ...

IT之家 12 月 2 日消息，The Verge 昨天采访到了 IBM 首席执行官 Arvind Krishna。他在播客表示，按照目前的数据中心建设与运营成本，行业投入的巨额资 本支出几乎不可能获得足够回报。 Arvind Krishna 指出，AI 企业在追求通用人工智能（AGI）的过程中不断扩大计算能力，但当前基础设施成本结构难以支撑这种规模化投资的经济可行性。 多位科技领域人士，包括 Marc Benioff、Andrew Ng 和 Mistral CEO Arthur Mensch，也对 AGI 的加速发展持保留意见。OpenAI 联合创始人 Ilya Sutskever 认 为大模型时代的"扩规模效应"已趋于极限，未来将重新进入以研究驱动的阶段。 尽管对 AGI 路线有所质疑，Krishna 仍肯定当前 AI 工具对企业生产力的价值，并认为这些技术将在企业领域释放"数万亿美元级"的效率收益。他提出未来 可能需要将硬知识体系与大模型结合，才能推动通用人工智能发展的下一步，但对其成功可能性仍保持慎重态度。 Krishna 表示，基于"今天的成本"进行的估算显示，一个 1 吉瓦的数据中心需要约 800 ...

Core Insights - The AI industry is transitioning from a "Scaling Era" back to a "Research Era," emphasizing fundamental innovation over mere model size expansion [4][7][40]. - Current AI models exhibit high performance in evaluations but lack true generalization capabilities, akin to students who excel in tests without deep understanding [10][25]. - SSI's strategy focuses on developing safe superintelligence without commercial pressures, aiming for a more profound understanding of AI's alignment with human values [15][16]. Group 1: Transition from Scaling to Research - The period from 2012 to 2020 was characterized as a "Research Era," while 2020 to 2025 is seen as a "Scaling Era," with a return to research now that computational power has significantly increased [4][7][40]. - Ilya Sutskever argues that simply scaling models will not yield further breakthroughs, as the data and resources are finite, necessitating new learning paradigms [7][39]. Group 2: Limitations of Current Models - Current models are compared to students who have practiced extensively but lack the intuitive understanding of true experts, leading to poor performance in novel situations [10][25]. - The reliance on pre-training and reinforcement learning has resulted in models that excel in benchmarks but struggle with real-world complexities, often introducing new errors while attempting to fix existing ones [20][21]. Group 3: Pursuit of Superintelligence - SSI aims to avoid the "rat race" of commercial competition, focusing instead on building a safe superintelligence that can care for sentient life [15][16]. - Ilya emphasizes the importance of a value function in AI, akin to human emotions, which guides decision-making and learning efficiency [32][35]. Group 4: Future Directions and Economic Impact - The future of AI is predicted to be marked by explosive economic growth once continuous learning challenges are overcome, leading to a diverse ecosystem of specialized AI companies [16][18]. - Ilya suggests that human roles may evolve to integrate with AI, maintaining balance in a world dominated by superintelligent systems [16][18].

Group 1 - The era of scaling is over, and the focus has shifted to research, emphasizing the importance of model generalization over mere computational power [4][8][34] - Emotional value functions are expected to play a crucial role in future AI developments, enhancing the efficiency of reinforcement learning [10][14][18] - The generalization ability of current models is still significantly inferior to that of humans, raising fundamental questions about AI's learning capabilities [13][19][25] Group 2 - The current models exhibit a "zigzag" capability, performing well in evaluations but struggling with real-world applications, indicating a disconnect between training and practical performance [27][30] - Companies that continue to pursue a scaling strategy may generate substantial revenue but could face challenges in achieving profitability due to intense competition [34][35] - The deployment of AI on a large scale could potentially lead to rapid economic growth, although the exact pace of this growth remains uncertain [35] Group 3 - Good research taste is essential, requiring a multi-faceted approach to identify beauty and simplicity in AI development [36][38] - The ultimate goal for AI development should be to create systems that genuinely care for and perceive life, rather than merely focusing on self-evolving AI [39][43] - The timeline for achieving superintelligence is projected to be within the next 5 to 20 years, contingent on advancements in understanding reliable generalization [44][46] Group 4 - SSI's current focus is on research, with plans to gradually deploy AI while ensuring that the first products released are meaningful and impactful [50][56] - SSI differentiates itself through a unique technical approach, aiming to create AI that is aligned with human values and capable of meaningful interaction [58]

Core Insights - The discussion centers on the limitations of current AI models and the new pathways toward superintelligence, emphasizing the disconnect between model performance in evaluations and real-world applications [3][4][20] - Ilya Sutskever highlights the need to transition back to a research-focused paradigm, moving away from mere scaling of models, as the diminishing returns of scaling become evident [3][34] - The concept of a "value function" is introduced as a critical element that enables human-like learning efficiency, which current AI lacks [3][5][6] Group 1: Current AI Limitations - Current AI models perform well in evaluation tests but often make basic errors in practical applications, indicating a lack of true understanding and generalization [4][18][20] - The over-optimization of reinforcement learning (RL) for evaluations has led to models that excel in competitive programming but struggle with real-world problem-solving [4][21] - Sutskever compares AI models to competitive programmers who are skilled in solving specific problems but lack the broader intuition and creativity of more versatile learners [4][22] Group 2: Human Learning Insights - Human learning is characterized by high sample efficiency, allowing individuals to learn complex skills with minimal data, attributed to innate value functions that guide decision-making [5][6][40] - The evolutionary advantages in human learning, particularly in areas like vision and motor skills, suggest that humans possess superior learning algorithms compared to current AI systems [5][38] - The discussion emphasizes the importance of emotional and intuitive feedback in human learning, which AI currently lacks [6][30][31] Group 3: Strategic Directions for SSI - Ilya Sutskever's new company, SSI, aims to explore safe superintelligence, advocating for a gradual release of AI capabilities to raise public awareness about safety [7][52] - The shift from a secretive development approach to a more transparent, gradual release strategy is seen as essential for fostering a collaborative safety environment [7][52] - SSI's focus on research over immediate market competition is intended to prioritize safety and ethical considerations in AI development [52][54] Group 4: Research Paradigm Shift - The transition from an era of scaling (2020-2025) back to a research-focused approach is necessary as the limits of scaling become apparent [34][46] - Sutskever argues that while scaling has been beneficial, it has also led to a homogenization of ideas, necessitating a return to innovative research [34][46] - The need for a more efficient use of computational resources in research is highlighted, suggesting that breakthroughs may come from novel approaches rather than sheer scale [35][46]

Core Insights - The AI industry is facing unprecedented bottlenecks as large model parameters reach trillion-level, with issues such as low efficiency of Transformer architecture, high computational costs, and disconnection from the physical world becoming increasingly prominent [2][4][38] - ZTE's recent paper, "Insights into Next-Generation AI Large Model Computing Paradigms," analyzes the core dilemmas of current AI development and outlines potential exploratory directions for the industry [2][38] Current State and Bottlenecks of LLMs - The performance of large language models (LLMs) is heavily dependent on the scaling laws, which indicate that ultimate performance is tied to computational power, parameter count, and training data volume [4][5] - Building advanced foundational models requires substantial computational resources and vast amounts of training data, leading to high sunk costs in the training process [5][6] - The efficiency of the Transformer architecture is low, with significant memory access demands, and the current hardware struggles with parallel operations in specific non-linear functions [6][7] Challenges in Achieving AGI - Current LLMs exhibit issues such as hallucinations and poor interpretability, which are often masked by the increasing capabilities driven by scaling laws [9][10] - There is ongoing debate regarding the ability of existing LLMs to truly understand the physical world, with criticisms focusing on their reliance on "brute force scaling" and lack of intrinsic learning and decision-making capabilities [9][10] Engineering Improvements and Optimizations - Various algorithmic and hardware improvements are being explored to enhance the efficiency of self-regressive LLMs, including attention mechanism optimizations and low-precision quantization techniques [12][13][14] - Innovations in cluster systems and distributed computing paradigms are being implemented to accelerate training and inference processes for large models [16][17] Future Directions in AI Model Development - The industry is exploring next-generation AI models that move beyond the Next-Token Prediction paradigm, focusing on models based on physical first principles and energy dynamics [24][26] - New computing paradigms, such as optical computing, quantum computing, and electromagnetic computing, are being investigated to overcome traditional computational limitations [29][30] ZTE's Exploration and Practices - ZTE is innovating at the micro-architecture level, utilizing advanced technologies to enhance AI accelerator efficiency and exploring new algorithms based on physical first principles [36][38] - The company is also focusing on the integration of hardware and software to create more efficient AI systems, contributing to the industry's shift towards sustainable development [38]