Market Overview - The gaming sector increased by 1.05% on January 6, with Caesar Culture leading the gains [1] - The Shanghai Composite Index closed at 4083.67, up 1.5%, while the Shenzhen Component Index closed at 14022.55, up 1.4% [1] Individual Stock Performance - Caesar Culture (002425) closed at 4.00, up 9.89% with a trading volume of 401,600 shares and a turnover of 158 million yuan [1] - Kaiying Network (002517) closed at 24.36, up 4.46% with a trading volume of 912,300 shares and a turnover of 2.179 billion yuan [1] - Gigabit (603444) closed at 448.99, up 2.77% with a trading volume of 24,400 shares and a turnover of 1.087 billion yuan [1] - Kunlun Wanwei (300418) closed at 44.89, up 2.56% with a trading volume of 949,700 shares and a turnover of 4.236 billion yuan [1] - Other notable stocks include Mingchen Health (002919) up 2.10%, Xunyou Technology (300467) up 2.04%, and Giant Network (002558) up 1.84% [1] Capital Flow Analysis - The gaming sector experienced a net outflow of 1.405 billion yuan from institutional investors, while retail investors saw a net inflow of 1.32 billion yuan [2] - The overall net inflow from speculative funds was 84.958 million yuan [2] Detailed Capital Flow for Selected Stocks - Kaiying Network (002517) had a net inflow of 255 million yuan from institutional investors, but a net outflow of 35.72 million yuan from speculative funds [3] - Caesar Culture (002425) saw a net inflow of 60.768 million yuan from institutional investors, with a net outflow of 36.423 million yuan from speculative funds [3] - Giant Network (002558) had a net inflow of 59.820 million yuan from institutional investors, but a significant net outflow of 82.100 million yuan from retail investors [3] - Other stocks like Xunyou Technology (300467) and Mingchen Health (002919) also showed varied capital flows, indicating mixed investor sentiment [3]

证券日报网讯 1月5日，昆仑万维发布公告称，公司于2025年12月19日召开的第五届董事会第三十六次 会议以及2026年1月5日召开的2026年第一次临时股东会，审议通过了《关于补选董事的议案》，同意选 举周亚辉先生担任公司第五届董事会董事，任期自2026年第一次临时股东会审议通过之日起至第五届董 事会任期届满之日止。 （文章来源：证券日报） ...

Core Insights - The article presents the 17th edition of the AI Product Rankings, focusing on the current state of AI applications as of December 2025, analyzing them through the lenses of scale, growth, and structure [9]. Global Rankings - ChatGPT remains the dominant AI application globally with 884.96 million MAU, showing a month-over-month growth of 14.1% [11][48]. - The top five AI applications in the global rankings also include four Chinese products: Doubao, Quark, Baidu Wangpan, and DeepSeek, with Doubao having 163.16 million MAU [12][48]. - Despite the high number of Chinese applications, ChatGPT's MAU is approximately five times that of Doubao, indicating a lack of a comparable "super entry" in the Chinese market [13][14]. Domestic Rankings - In the domestic market, the top five applications are Doubao (163.16 million MAU), Quark (147.84 million MAU), Baidu Wangpan (142.06 million MAU), DeepSeek (131.48 million MAU), and Tencent Yuanbao (84.38 million MAU) [15][36]. - The commonality among these top applications is their origin from major tech companies, highlighting the importance of established platforms in capturing high-frequency usage scenarios [16][17]. Growth Rankings - The most notable growth in global MAU was seen in xMate, which increased by 44 times, followed by Music AI with a 16-fold increase [18][19]. - In the domestic market, DeepSeek achieved a remarkable 89.53% growth in MAU, indicating its strong structural significance [23]. International Expansion - ByteDance's Doubao AI assistant, now rebranded as Dola, reached 53.63 million MAU with a growth of 4.72% [25][39]. - Many products in the international market focus on specific use cases, achieving significant breakthroughs in areas like video generation and efficiency tools [25][26]. User Behavior Insights - Two critical questions determine the sustainability of AI products: whether users are experimenting or forming habits, and whether growth is driven by marketing or the product itself [28]. - Products that consistently appear in both new user download rankings and MAU growth rankings are likely to have found product-market fit (PMF) [28]. Subscription Revenue Rankings - The subscription revenue rankings indicate that ChatGPT leads with an annualized revenue of $198.48 million, followed by Grok and Claude with $4.71 million and $2.98 million respectively [46][47].

北京市天元律师事务所 关于昆仑万维科技股份有限公司 2026 年第一次临时股东会的法律意见 京天股字（2026）第 008 号 致：昆仑万维科技股份有限公司 昆仑万维科技股份有限公司（以下简称"公司"）2026 年第一次临时股东会（以 下简称"本次股东会"）采取现场投票与网络投票相结合的方式，现场会议于 2026 年 1 月 5 日在北京市东城区西总布胡同 46 号明阳国际中心 B 座 11 层会议室召开。 北京市天元律师事务所（以下简称"本所"）接受公司聘任，指派本所律师参加本次 股东会现场会议，并根据《中华人民共和国公司法》《中华人民共和国证券法》（以 下简称"《证券法》"）、《上市公司股东会规则》（以下简称"《股东会规则》"） 以及《昆仑万维科技股份有限公司章程》（以下简称"《公司章程》"）等有关规定， 就本次股东会的召集、召开程序、出席现场会议人员的资格、召集人资格、会议表 决程序及表决结果等事项出具本法律意见。 本所及经办律师依据《证券法》《律师事务所从事证券法律业务管理办法》和《律 师事务所证券法律业务执业规则（试行）》等规定及本法律意见出具日以前已经发生 或者存在的事实，严格履行了法定职责，遵循 ...

证券代码：300418 证券简称：昆仑万维 公告编号：2026-001 昆仑万维科技股份有限公司 2026 年第一次临时股东会决议公告 本公司及董事会全体成员保证信息披露的内容真实、准确、完整，没有虚假记载、 误导性陈述或重大遗漏。 重要提示： 1、召集人：昆仑万维科技股份有限公司董事会 1、本次股东会没有否决或修改提案的情况。 2、为尊重中小投资者利益，提高中小投资者对公司股东会决议事项的参与度，本 次股东会对中小投资者进行单独计票，中小投资者是指除上市公司董事、高级管理人员 以及单独或者合计持有公司 5%以上股份的股东以外的其他股东。 3、本次股东会以现场与网络投票相结合的方式召开。 一、会议通知情况 昆仑万维科技股份有限公司（以下简称"公司"）2026 年第一次临时股东会通知的 公告（公告编号：2025-087）已于 2025 年 12 月 20 日刊登于中国证监会指定的创业板 信息披露媒体巨潮资讯网（www.cninfo.com.cn）。 二、会议召开和出席情况 （一）会议召开情况 2、主持人：公司董事长方汉先生 （2）网络投票时间： 2026 年 1 月 5 日，其中，通过深交所交易系统进行网络投票 ...

证券代码：300418 证券简称：昆仑万维 公告编号：2026-002 昆仑万维科技股份有限公司 关于完成补选董事的公告 本公司及董事会全体成员保证信息披露的内容真实、准确、完整，没有虚假记载、误导性 陈述或重大遗漏。 昆仑万维科技股份有限公司（以下简称"公司"）于2025年12月20日披露了《关于董事辞 职及补选董事的公告》（2025-085）。董事黄国强先生因个人原因，申请辞去公司第五届董事 会董事、第五届董事会审计委员会委员职务，辞职后不再在公司担任任何职务。 公司于2025年12月19日召开的第五届董事会第三十六次会议以及2026年1月5日召开的 2026年第一次临时股东会，审议通过了《关于补选董事的议案》，同意选举周亚辉先生担任公 司第五届董事会董事，任期自2026年第一次临时股东会审议通过之日起至第五届董事会任期届 满之日止。此次补选完成后，董事会中兼任公司高级管理人员以及由职工代表担任的董事人数 总计不超过公司董事总数的二分之一。 特此公告。 昆仑万维科技股份有限公司董事会 二零二六年一月五日 ...

Group 1 - The core point of the news is that Kunlun Wanwei's stock price increased by 5.23% to 43.88 CNY per share, with a trading volume of 3 billion CNY and a market capitalization of 55.085 billion CNY as of January 5 [1] - Kunlun Wanwei, established on March 27, 2008, and listed on January 21, 2015, is based in Beijing and operates in the fields of comprehensive internet value-added services and new energy investment [1] - The company's revenue composition includes: 38.37% from online advertising, 18.51% from Opera search, 15.61% from short drama platforms, 13.92% from overseas social networks, 6.40% from gaming, 4.27% from social entertainment platforms, 1.75% from AI software technology, and 1.16% from other businesses [1] Group 2 - From the perspective of major fund holdings, one fund under Nuoan Fund has a significant position in Kunlun Wanwei, with Nuoan Active Return Mixed A (001706) reducing its holdings by 1.2285 million shares to 2.2285 million shares, representing 4.82% of the fund's net value [2] - The fund has generated an estimated floating profit of approximately 4.8581 million CNY today [2] - Nuoan Active Return Mixed A (001706) was established on September 22, 2016, with a current scale of 1.365 billion CNY, and has experienced a loss of 4.63% this year, ranking 8061 out of 8155 in its category [2]

Core Insights - The hybrid expert model (MoE) is recognized as a "structural revolution" in artificial intelligence, enabling the construction of ultra-large-scale and high-efficiency models through its sparse activation design [1][7] - The market size for China's MoE industry is projected to reach approximately 148 million yuan in 2024, reflecting a year-on-year growth of 43.69% [1][7] - The sparse activation mechanism allows models to scale to trillions of parameters at a significantly lower computational cost compared to traditional dense models, achieving a revolutionary balance between performance, efficiency, and cost [1][7] Industry Overview - MoE is a neural network architecture that enhances performance and efficiency by dynamically integrating multiple specialized sub-models (experts), focusing on a "divide-and-conquer strategy + conditional computation" [2][3] - The core characteristics of MoE include high parameter capacity and low computational cost, activating only a small portion of total parameters to expand model size [2][3] - MoE faces technical challenges such as load balancing, communication overhead among experts, and high memory requirements, while offering advantages like task specificity, flexibility, and efficiency [2][3] Industry Development History - The MoE concept originated from the "adaptive mixture of local experts" theory proposed by Michael Jordan and Geoffrey Hinton in 1991, focusing on efficient collaboration through a gating network [3][4] - Significant advancements occurred in 2017 when Google introduced sparse gating mechanisms in LSTM networks, leading to substantial reductions in computational costs and performance breakthroughs in NLP tasks [3][4] - The MoE technology has rapidly evolved alongside deep learning and big data trends, with notable models like Mistral AI's Mixtral 8x7B and DeepSeek-MoE series pushing the boundaries of performance and efficiency [3][4] Industry Value Chain - The upstream of the MoE industry includes chips, storage media, network devices, and software tools for instruction sets and communication libraries [6] - The midstream focuses on the development and optimization of MoE models, while the downstream applications span natural language processing, computer vision, multimodal large models, and embodied intelligence [6] - The natural language processing market in China is expected to reach approximately 12.6 billion yuan in 2024, growing by 14.55% year-on-year, driven by technological breakthroughs and increasing demand across various sectors [6] Market Size - The MoE industry in China is projected to reach a market size of about 148 million yuan in 2024, with a year-on-year growth rate of 43.69% [1][7] - The technology's advantages are attracting significant investments from research institutions, large tech companies, and AI startups, facilitating the transition from technical prototypes to scalable commercial applications [1][7] Key Company Performance - The MoE industry in China is characterized by a competitive landscape involving "open-source pioneers, large enterprises, and vertical deep-divers," with market concentration undergoing dynamic reshaping [8][9] - Leading companies like Kunlun Wanwei and Tencent are leveraging technological innovation and product advantages to establish a strong market position [8][9] - Kunlun Wanwei launched the first domestic open-source model based on MoE architecture in February 2024, achieving a threefold increase in inference efficiency compared to dense models [9] Industry Development Trends - The demand for multimodal data is driving the integration of MoE architecture with technologies like computer vision and speech recognition, making multimodal MoE models mainstream [10] - Breakthroughs in sparse activation and expert load balancing technologies are enhancing the stability and inference efficiency of large-scale MoE models [11] - The construction of ecosystems around open-source frameworks and domestic computing power is accelerating the large-scale implementation of MoE in various fields [12]

Group 1 - The AI corpus concept rose by 2.00%, ranking 7th among concept sectors, with 39 stocks increasing in value, including DeSheng Technology and Zhangyue Technology hitting the daily limit up [1] - Notable gainers in the AI corpus sector included FuShi Holdings, ZhiDeMai, and Century Tianhong, which increased by 10.68%, 6.98%, and 5.54% respectively [1] - The sector saw a net inflow of 1.927 billion yuan from main funds, with 33 stocks receiving net inflows, and 6 stocks exceeding 100 million yuan in net inflow [2] Group 2 - The top net inflow stock was Kunlun Wanwei, with a net inflow of 394 million yuan, followed by Zhangyue Technology, Keda Xunfei, and DeSheng Technology with net inflows of 295 million yuan, 265 million yuan, and 150 million yuan respectively [2] - The net inflow ratios for DeSheng Technology, Zhangyue Technology, and Hanwang Technology were 35.28%, 25.56%, and 16.69% respectively, indicating strong investor interest [3] - The AI corpus sector's performance was contrasted with other sectors, such as the silicon energy sector, which saw a decline of 1.03% [2]

Market Overview - The gaming sector experienced a slight decline of 0.13% on December 31, with Century Huatong leading the drop [1] - The Shanghai Composite Index closed at 3968.84, up 0.09%, while the Shenzhen Component Index closed at 13525.02, down 0.58% [1] Individual Stock Performance - Kunlun Wanwei (300418) saw a closing price of 41.70, with an increase of 3.78% and a trading volume of 811,500 shares, amounting to a transaction value of 3.371 billion [1] - Xunyou Technology (300467) closed at 28.70, up 2.76%, with a trading volume of 73,300 shares [1] - Tom Cat (300459) closed at 4.59, up 2.68%, with a trading volume of 1,528,300 shares [1] - Other notable performers include Shenzhou Taiyue (300002) with a 1.77% increase and a closing price of 11.52, and Sheng Tian Network (300494) with a 1.50% increase and a closing price of 12.18 [1] Fund Flow Analysis - The gaming sector saw a net inflow of 375 million from institutional investors, while retail investors experienced a net outflow of 145 million [2] - Major stocks like Kunlun Wanwei had a net inflow of 413 million from institutional investors, while retail investors had a net outflow of 254 million [3] - Shenzhou Taiyue also reported a net inflow of 87.39 million from institutional investors, with retail investors seeing a net outflow of 52.79 million [3] ETF Performance - The Gaming ETF (product code: 159869) tracked the China Animation and Gaming Index, with a slight decline of 0.06% over the past five days and a price-to-earnings ratio of 37.03 [5] - The latest share count for the Gaming ETF is 8.77 billion, with an increase of 65 million shares and a net subscription of 930.77 million [5]