Core Viewpoint - The market is experiencing a significant shift towards the AI sector, with various companies in the AI application and industry chain showing strong performance and growth [1][2]. Group 1: Market Performance - AI-related stocks such as Jiecheng Co., Zhongwen Online, and Xingfu Lanhai have seen their shares hit the daily limit, with Jiecheng Co. rising by 20.03% and Zhongwen Online by 20.01% [2][3]. - Notable stock performances include: - Jiecheng Co. at 8.45, up by 1.41, with a market cap of 225.09 billion and a year-to-date increase of 53.08% [3]. - Zhongwen Online at 42.34, up by 7.06, with a market cap of 308.45 billion and a year-to-date increase of 68.55% [3]. - Happiness Blue Sea at 29.46, up by 4.91, with a market cap of 109.77 billion and a year-to-date increase of 41.70% [3]. - In the Hong Kong market, stocks like Zhiyu and Yuedu Group have also shown significant gains, with Zhiyu rising over 14% and Yuedu Group over 13% [2][4]. Group 2: AI Technology Development - The launch of Seedance 2.0, an AI video generation model, has been a pivotal moment in the AI application market, being described as a "director-level" full-process generation engine [5][6]. - Seedance 2.0 utilizes a "dual-branch diffusion transformer" architecture, capable of generating both video and audio simultaneously, achieving over 90% usability in practical applications [7][8]. - The model is expected to significantly reduce production costs for short dramas and indicates a shift towards industrialized production in AI video [9][10]. Group 3: Competitive Landscape - The global AI model competition is intensifying, with major players like Alibaba, Google, and others launching their models, marking the beginning of a "model war" [12][13]. - The emergence of various AI models, including Grok Imagine and Claude Opus, highlights the rapid advancements in AI technology and the competitive dynamics within the industry [13][14]. - Analysts suggest that the recent developments in AI video models are pushing the boundaries of technology in the video generation field, similar to the competition seen in large language models in 2025 [14][15].

春节前后是国内大厂密集发布AI应用的关键窗口。近日，字节跳动旗下即梦AI发布新一代视频生 成模型Seedance 2.0，实现多项质变，其多镜头叙事与导演级控制、精准的音画同步与口型匹配，以及 大幅增强的角色与场景一致性，共同使AI视频具备了长篇叙事能力，正朝影视工业级别迈进。 （责任编辑：张晓波 ） 在此之前，阿里云推出性能媲美国际顶尖水平的推理模型Qwen3-Max-Thinking，标志着国产模型 在高阶推理领域的重要突破；DeepSeek发布并开源能处理复杂图片的DeepSeek-OCR 2模型。多家企业 接连发布新品，一场围绕AI技术的春节"竞赛"正在上演。 港股通互联网ETF易方达（513040）跟踪的中证港股通互联网指数聚焦港股互联网平台公司，汇聚 中国AI核心企业；软件ETF易方达（562930）跟踪的中证软件服务指数则由A股中30只业务涉及软件开 发、软件服务等领域的股票组成，覆盖AI+办公、AI+金融、AI+教育等各类落地场景，可助力投资者高 效把握AI应用领域布局机会。 2月10日午后，AI应用方向持续活跃。截至14点，中证港股通互联网指数上涨0.3%，成分股中，阿 里巴巴-W、小米集团 ...

C 则联社 | AI应用"春节档"持续上新 "神仙打架"已就位? 2月9日,字节跳动旗下Seedance 2.0 Al视频生成模型 刷屏海内外互联网,点燃A股"AI应用"板块,引力传媒、 荣信文化、中文在线等多只概念股涨停。事实上,在整 个"春节档",字节跳动、阿里千问、快手、百度等多家 大厂已先后发布AI应用新产品，后续或将还有Qwen 35 DeenSeekV4 豆句20等新模型面世。 资料来源:公开报道以及研报等 为速度和交易而生 上海报业集团主管主办 编辑:牛嘉 制图:孙罕颖 AI应用"春节档"新品一览 | 公司 | 产品 | 产品 | 核心能力 | | --- | --- | --- | --- | | 字节跳动 | Seedance 2.0 | Al生成视频 | 可根据文本或图像创建 | | | (即梦) | | 电影级频频 | | 阿里千问 | Qwen3-Max- | Al助理 | 支持自适应工具调用和 | | | Thinking | | 测试时扩展技术 | | | 文心 Paddle | | 进一步集成印章识别、 | | 百度文心 | OCR-VL-1.5 | Al办公 | 文本检测与识别 ...

Investment Rating - The industry investment rating is "Overweight" [1] Core Insights - Alibaba's Qwen3-Max-Thinking model has achieved significant improvements in various key dimensions, including factual knowledge, complex reasoning, instruction adherence, human preference alignment, and agent capabilities, performing comparably to top models like GPT-5.2-Thinking and Claude-Opus-4.5 in 19 authoritative benchmark tests [3][12] - The model features over one trillion parameters and enhances the native agent capabilities for autonomous tool invocation, significantly reducing hallucinations and laying the groundwork for solving real complex tasks [5][12] - The release of Clawdbot marks a new phase in the development of intelligent agents, allowing users to interact through messaging apps and execute commands via local scripts, thus transitioning from "AI providing suggestions" to "AI taking direct actions" [4][13] Summary by Sections 1. Computer Industry Insights - On January 26, Alibaba released the flagship reasoning model Qwen3-Max-Thinking, which has shown significant enhancements across multiple key performance metrics [12] - The model's core innovations include adaptive tool invocation capabilities and test-time scaling technology, which significantly boosts reasoning performance [12][14] 2. Market Performance Review - The computer industry index fell by 4.77%, underperforming the Shanghai Composite Index by 4.33 percentage points, while the year-to-date performance shows a 6.94% increase [16][17] - The computer sector ranked 28th among 31 industry indices this week, indicating a challenging market environment [16] 3. Technology Software Industry News - The "Smart Computing" initiative in Shanghai aims to enhance the AI infrastructure, with a focus on building a complete industrial chain covering chips, optical communication, and servers [23] - The launch of the "Star Computing" plan by a commercial aerospace company aims to create a space-based computing network to support AI agents with low-latency and high-reliability services [24] 4. Company Dynamics - Companies like Cambricon, Inspur, and Alibaba are highlighted as key players in the AI industry chain, suggesting investment opportunities in these firms [5][14] - Recent stock option and incentive plans from various companies indicate a focus on retaining talent and driving growth in the tech sector [35][36]

2026 开年没多久，大模型圈就又要热到火起来了。 仅仅是 2 月前后这几天，Kimi 2.5、Qwen3-Max-Thinking、Step 3.5 Flash 就接连发布。Qwen3-Max-Thinking 直接对标 GPT-5.2-Thinking、Claude-Opus-4.5 和 Gemini 3 Pro 等旗舰推理模型，Kimi 2.5、Step 3.5 Flash 则瞄准了 Agent 时代的模型升级。 今天（2 月 6 日）凌晨，OpenAI、Anthropic 也几乎同时推出了小版本迭代的 GPT-5.3-Codex 和 Claude Opus 4.6。 但这一切还是刚开始，阿里 Qwen 研究员郑楚杰在 Qwen3-Max-Thinking 发布时强调这是 Qwen 迄今为止最好的模型，同时又透露「Qwen 3.5 的发布也指日 可待」。 如果把过去两年的大模型竞争总结为「更大、更强、更全」，那么 2026 年这一轮更新，方向已经明显变了。 图片来源：X 指日可待的不只是 Qwen 3.5。 1 月初智谱上市，智谱首席科学家唐杰就在内部信中透露即将推出新一代模型 GLM-5，最近南华早报 ...

炒股就看金麒麟分析师研报，权威，专业，及时，全面，助您挖掘潜力主题机会！ 来源：人工智能顾问 日前，国内权威大模型评测机构SuperCLUE正式发布《2025年年度中文大模型基准测评报告》。这份被 誉为"AI界年终成绩单"的重磅榜单，再次点燃了行业关注。 这标志着国产大模型正从"追赶者"迈向"并行者"，甚至在特定领域实现"领跑"。 01 海外闭源模型领跑，国产模型紧咬不放 本次测评从数学推理、科学理解、代码生成等六大核心维度，全面检验大模型的综合智能水平。最终综 合得分排名中： ●第1名：Claude-Opus-4.5-Reasoning（Anthropic）—68.25分 ●第2名：Gemini-3-Pro-Preview（谷歌）—65.59分 ●第3名：GPT-5.2（high）（OpenAI）—64.32分 三大海外闭源模型凭借其强大的工程优化与数据积累，继续领跑全球。 但值得注意的是，第4名由国产开源模型Kimi-K2.5-Thinking以61.50分强势拿下，成为综合排名最高的 国产模型；而阿里云推出的高性能模型Qwen3-Max-Thinking则以60.61分位列第6，紧随其后，展现出中 国头 ...

Core Viewpoint - The major internet companies Tencent, Alibaba, Baidu, and ByteDance are engaged in a "red envelope war" as they prepare for the 2026 Spring Festival, with significant cash giveaways aimed at gaining user engagement and establishing dominance in the AI sector [2][6]. Group 1: Cash Red Envelope Initiatives - Tencent announced a 10 billion yuan cash red envelope campaign starting February 1, 2026, while Baidu launched a 5 billion yuan initiative on the same day [2][4]. - Alibaba's Qianwen APP revealed a massive 30 billion yuan "Spring Festival Treat Plan" on February 2, 2026, marking its largest investment in a Spring Festival campaign [4]. - Users can participate in Tencent's red envelope activities by logging in daily, completing tasks, and sharing links, with the potential to win significant cash prizes [3][4]. Group 2: AI Integration and User Engagement - The red envelope campaigns are seen as a strategy to test user engagement with AI applications, marking a shift from previous marketing tactics that focused solely on cash giveaways [6][9]. - Industry experts suggest that the effectiveness of these campaigns will depend on the quality and usability of the AI products offered, rather than just the cash incentives [9][10]. - The competition is characterized by each company's focus on integrating their core business strengths, with Tencent leveraging social features, Alibaba focusing on a comprehensive ecosystem, and Baidu enhancing search functionalities [5][6]. Group 3: Long-term Implications for AI Market - The ongoing "red envelope war" is viewed as a significant event in the AI landscape, potentially accelerating the adoption and application of AI technologies across various sectors [8]. - Analysts predict that the competition will lead to a more profound integration of AI into everyday life, with companies needing to address real user needs to ensure long-term retention [10]. - The AI market in China is expected to grow significantly, with projections indicating a core industry scale surpassing 900 billion yuan by 2024, reflecting a 24% year-on-year increase [10].

Core Viewpoint - The release of the Step3.5Flash model by Jiyue Xingchen marks a significant advancement in the domestic AI landscape, transitioning from a focus on larger models to smarter, more efficient models capable of real-world applications [1][2]. Group 1: Model Capabilities - The Step3.5Flash model boasts a reasoning speed of up to 350 TPS (tokens per second) and utilizes a sparse MoE (Mixture of Experts) architecture with 196 billion parameters, making it a leading open-source model for agent scenarios in China [1][2]. - It can efficiently handle 256K context, making it suitable for complex tasks with long logical chains, and its performance in mathematical reasoning and coding tasks is competitive with mainstream closed-source models [1][2]. Group 2: Practical Applications - The model can perform complex mathematical calculations and automate programming tasks based on text prompts, showcasing its practical utility [2]. - In a demonstration, Step3.5Flash effectively compared prices across major e-commerce platforms, illustrating its ability to break down complex tasks into manageable subtasks and enhance user experience [3]. Group 3: Industry Context - The release coincides with a competitive landscape where major companies like Alibaba and Moonlight are also advancing their agent capabilities, indicating a race for dominance in the AI agent sector [5][6]. - Jiyue Xingchen's strategy emphasizes efficiency and cost-effectiveness in deploying agents, aiming to lower the barriers for developers to implement high-performance agents on consumer devices [6][7]. Group 4: Future Directions - The company is shifting its focus towards mobile, PC, and smart cockpit applications, with expectations that Step3.5Flash will be integrated into customized terminals through a cloud-edge collaboration model [7]. - The founder of Jiyue Xingchen highlighted the importance of multimodal perception and long-chain reasoning as key drivers for the evolution of intelligent agents, suggesting that the new model could significantly empower the company's future agent development [7].

Core Insights - The Chinese large model sector is experiencing accelerated development in 2025, with the SuperCLUE annual evaluation covering 23 representative models from both domestic and international sources, focusing on general capabilities, specialized tasks, and application scenarios [1][2]. Group 1: Model Performance - The top-ranking closed-source model is Anthropic's Claude-Opus-4.5-Reasoning, scoring 68.25, followed by Google Gemini-3-Pro-Preview and OpenAI GPT-5.2 (high) [1][23]. - Domestic models are transitioning from "catching up" to "running alongside," with Kimi-K2.5-Thinking (61.50) and Qwen3-Max-Thinking (60.61) ranking fourth and sixth globally, excelling in code generation and mathematical reasoning tasks [1][2][23]. - The performance gap in precise instruction adherence and hallucination control remains significant, with average score differences exceeding 7 points and nearly 2 points, respectively [2]. Group 2: Technological Evolution - The evolution of technology is characterized by three stages: early competition among numerous models and the emergence of multimodal capabilities, a mid-stage explosion of multimodal applications and reasoning breakthroughs, and the rise of intelligent agents and ecosystem reconstruction by 2025 [1][2]. - The mixed expert (MoE) architecture has become mainstream, with domestic open-source models capturing a significant share of the global market, led by DeepSeek and Qwen3 [1][2]. Group 3: Application and Cost-Effectiveness - In application scenarios, general intelligent agents are still in their foundational stages, lacking in complex task handling capabilities; however, domestic models excel in multimodal areas such as image-to-video generation and Chinese adaptation [2]. - Domestic models demonstrate significant cost-effectiveness, with Kimi-K2.5-Thinking priced at only one-third of similar overseas models, although overseas models outperform in reasoning efficiency [2]. Group 4: Future Directions - The Chinese large model sector has made significant advancements in technological innovation, application deployment, and ecosystem construction, establishing core competitive advantages in open-source ecosystems, vertical applications, and cost-effectiveness [2]. - Future efforts should focus on overcoming shortcomings in precise instruction adherence and hallucination control to drive technology towards more efficient and reliable outcomes [2].

Investment Rating - The report does not explicitly provide an investment rating for the industry or companies involved. Core Insights - The report highlights significant advancements in Chinese large models and AI agents, marking a transition from "following" to "keeping pace" with global leaders in AI technology [14][24]. - The competitive landscape shows a clear distinction between domestic and international models, with domestic open-source models gaining substantial ground [23][47]. - The report emphasizes the importance of multi-modal capabilities and the emergence of AI agents in practical applications, particularly in programming and task planning [16][14]. Summary by Sections 1. Key Developments in 2025 - The report outlines three major phases of AI model evolution: the initial competition among models, the explosion of multi-modal capabilities, and the rise of AI agents [14][16]. - Notable models such as Kimi-K2.5-Thinking and Qwen3-Max-Thinking have emerged as leaders in specific tasks like code generation and mathematical reasoning [18][24]. 2. Annual Evaluation Results and Analysis - The 2025 annual evaluation ranks Claude-Opus-4.5-Reasoning as the top model globally, followed by Gemini-3-Pro-Preview and GPT-5.2(high) [23][45]. - Domestic models like Kimi-K2.5-Thinking and Qwen3-Max-Thinking are positioned fourth and sixth, indicating a strong competitive stance [23][45]. - The report notes that domestic models are rapidly closing the gap with international counterparts, particularly in code generation and reasoning tasks [24][48]. 3. SuperCLUE Model Quadrant and Capability Landscape - The report presents a model quadrant that categorizes models based on their capabilities in reasoning and application, highlighting the emergence of "technical leaders" and "practical leaders" in the domestic market [38][39]. - The capability landscape indicates that while domestic models excel in certain areas, they still face challenges in hallucination control and precise instruction adherence [42][48]. 4. Comparative Analysis of Domestic and International Models - The analysis reveals that closed-source models dominate the top rankings, with significant advantages in reasoning and instruction-following tasks [74][80]. - Domestic open-source models are noted for their rapid advancements, particularly in coding tasks, where they have begun to outperform some international models [56][84]. - The report emphasizes the structural differences between domestic and international models, with domestic models showing a strong trend towards open-source development [24][47].