每经AI快讯，1月26日，阿里千问官微宣布正式推出其最新旗舰推理模型Qwen3-Max-Thinking。据介 绍，Qwen3-Max-Thinking在多个关键维度上实现了显著提升，包括事实知识、复杂推理、指令遵循、 人类偏好对齐以及智能体能力。在19项权威基准测试中，其性能可媲美GPT-5.2-Thinking、Claude-Opus- 4.5和Gemini3Pro等顶尖模型。 ...

Core Viewpoint - The article highlights the launch of Qwen3-Max-Thinking by Alibaba Qwen, which has achieved state-of-the-art (SOTA) performance in various benchmark tests, surpassing leading models like GPT-5.2-Thinking and Claude-Opus-4.5 in multiple categories [1][2]. Group 1: Model Performance - Qwen3-Max-Thinking has demonstrated superior performance in 19 authoritative benchmark tests, achieving scores that match or exceed those of top closed-source models [1]. - In the MMLU-Pro benchmark, Qwen3-Max-Thinking scored 85.7, while GPT-5.2-Thinking scored 87.4, and Claude-Opus-4.5 scored 89.5 [2]. - The model's reasoning capabilities were highlighted, achieving a score of 91.5 in the IMO-AnswerBench, the highest among competitors [31]. Group 2: Technical Innovations - Qwen3-Max-Thinking incorporates two key innovations: adaptive tool invocation and test-time scaling, which significantly enhance its reasoning performance and native agent capabilities [3][19]. - The adaptive tool invocation allows the model to autonomously select and utilize built-in functions such as search and code interpreters during interactions, improving efficiency [22][24]. - Test-time scaling allocates additional computational resources during the reasoning phase, leading to improved performance without unnecessary redundancy [27][30]. Group 3: Market Impact and Adoption - The article notes that Chinese open-source AI models have gained significant traction, with a 17.1% adoption rate in global model downloads, surpassing the U.S. at 15.8% [36]. - Alibaba's Qwen series has achieved over 10 billion downloads, averaging 1.1 million downloads per day, establishing itself as a new benchmark in the global AI open-source community [39]. - The integration of Qwen models into Alibaba's ecosystem, including platforms like Taobao and Alipay, indicates a strategic focus on combining top-tier model capabilities with practical applications [42][43].

Core Insights - Alibaba officially launched the Qwen3-Max-Thinking flagship reasoning model on January 26, setting multiple global records in authoritative evaluations, with performance comparable to GPT-5.2 and Gemini 3 Pro, making it the strongest domestic AI large model closest to international top models [1] Group 1 - The new Qwen model has over one trillion parameters and underwent extensive reinforcement learning training, achieving a significant leap in model performance through a series of innovations in reasoning technology [1] - Hugging Face's latest data shows that the number of Qwen derivative models has surpassed 200,000, making it the first open-source large model to reach this milestone globally [1] - The Qwen series models have exceeded 1 billion downloads, averaging 1.1 million downloads per day by developers, maintaining the top position in global open-source large models [1]

Group 1: Core Insights - The report highlights a decline in trading heat for hot themes, with commercial aerospace, resource products, and building materials leading the gains, while light communication and consumer electronics have seen a pullback [1][7] - The focus of incremental policies is on urban renewal and emerging technologies, with a positive outlook on urban renewal, commercial aerospace, domestic computing power, and new power grids [1][6] Group 2: Urban Renewal - Urban renewal is expected to stabilize the real estate market and expand domestic demand, with a projected 60,015 urban renewal projects nationwide in 2024, totaling an investment of 2.9 trillion yuan [19][25] - Key areas of urban renewal include the renovation of old residential communities, urban villages, and underground pipeline upgrades, with recommendations for investment in waterproofing, piping, and coatings [20][26] Group 3: Commercial Aerospace - Elon Musk indicated that space will become the lowest-cost location for deploying AI data centers within 2-3 years, with reusable spacecraft potentially reducing access costs by 100 times [21][36] - The report anticipates the establishment of a leading space computing center in China by 2030, focusing on green and low-carbon solutions [38][40] Group 4: Domestic Computing Power - TSMC's capital expenditure is expected to exceed expectations, with a projected 37% increase in 2026 to reach $56 billion, indicating strong demand in the semiconductor manufacturing sector [21][48] - The rapid iteration of domestic AI models is expected to drive investment demand in domestic computing power, with significant increases in user engagement and data usage [22][49] Group 5: New Power Grids - The State Grid's investment during the 14th Five-Year Plan is projected to reach 4 trillion yuan, a 40% increase from the previous plan, aimed at accelerating the construction of new power systems [21][22] - By 2025, non-fossil energy consumption in China is expected to account for 20% of total energy consumption, increasing to 25% by 2030 [21][22]

格隆汇1月26日丨国际医学(000516.SZ)在投资者互动平台表示，公司目前与阿里巴巴暂不涉及AI诊断方 面的合作。 ...

成交活跃股方面，今日上榜个股中，南向资金成交金额最多的是阿里巴巴-W，合计成交额56.22亿港 元，腾讯控股、中国海洋石油成交额紧随其后，分别成交43.07亿港元、33.93亿港元。以净买卖金额统 计，净买入的个股共有5只，腾讯控股净买入额为10.15亿港元，净买入金额居首，该股收盘股价上涨 0.76%，小米集团-W净买入额为8.22亿港元，泡泡玛特净买入额为6.13亿港元。净卖出金额最多的是中 国移动，净卖出11.77亿港元，该股收盘股价下跌0.88%，紫金矿业、盈富基金遭净卖出10.77亿港元、 10.36亿港元。 今日上榜个股中，腾讯控股、小米集团-W、钧达股份等7只股同时上榜港股通（深）、港股通（沪）成 交活跃股， 腾讯控股合计成交额43.07亿港元，成交净买入10.15亿港元，小米集团-W合计成交额30.90 亿港元，成交净买入8.22亿港元。紫金矿业合计成交额33.65亿港元，成交净卖出10.77亿港元，阿里巴 巴-W合计成交额56.22亿港元，成交净卖出2.35亿港元。 1月26日恒生指数上涨0.06%，南向资金全天合计成交金额为1145.35亿港元，其中，买入成交568.55亿 港元，卖出成交 ...

| 代码 | 简称 | 类型 | 成交金额（万港元） | 成交净买入（万港元） | 日涨跌幅（%） | | --- | --- | --- | --- | --- | --- | | 09988 | 阿里巴巴-W | 港股通(沪) | 374381.67 | -12829.26 | -1.96 | | 00700 | 腾讯控股 | 港股通(沪) | 256404.25 | 47186.91 | 0.76 | | 02899 | 紫金矿业 | 港股通(沪) | 234866.76 | -160485.92 | 4.35 | | 01810 | 小米集团-W | 港股通(沪) | 202530.35 | 52716.55 | -2.81 | | 02865 | 钧达股份 | 港股通(沪) | 200947.57 | 22193.94 | -6.56 | | 00941 | 中国移动 | 港股通(沪) | 199564.94 | -117692.48 | -0.88 | | 00981 | 中芯国际 | 港股通(沪) | 197489.33 | 18826.54 | -3.27 | | 09988 | 阿里巴巴-W ...

Core Insights - On January 26, 2026, Alibaba-W (09988), Tencent Holdings (00700), and Zijin Mining (02899) were the top three companies by trading volume in the Southbound Stock Connect, with trading amounts of 3.744 billion, 2.564 billion, and 2.349 billion respectively [1] - In the Shenzhen-Hong Kong Stock Connect, Alibaba-W (09988), Tencent Holdings (00700), and China National Offshore Oil Corporation (00883) also ranked as the top three, with trading amounts of 1.878 billion, 1.743 billion, and 1.608 billion respectively [1] Southbound Stock Connect - Top Active Companies - Alibaba-W (09988) had a trading amount of 3.744 billion with a net buy of -128 million [2] - Tencent Holdings (00700) recorded a trading amount of 2.564 billion with a net buy of 472 million [2] - Zijin Mining (02899) had a trading amount of 2.349 billion with a net buy of -1.605 billion [2] - Xiaomi Group-W (01810) achieved a trading amount of 2.025 billion with a net buy of 527 million [2] - JunDa Co., Ltd. (02865) had a trading amount of 2.009 billion with a net buy of 222 million [2] - China Mobile (00941) recorded a trading amount of 1.996 billion with a net buy of -1.177 billion [2] - SMIC (00981) had a trading amount of 1.975 billion with a net buy of 188 million [2] - China National Offshore Oil Corporation (00883) achieved a trading amount of 1.785 billion with a net buy of -35.021 million [2] - Pop Mart (09992) had a trading amount of 1.419 billion with a net buy of 613 million [2] - Yangtze Optical Fibre and Cable (06869) recorded a trading amount of 1.371 billion with a net buy of -6.072 million [2] Shenzhen-Hong Kong Stock Connect - Top Active Companies - Alibaba-W (09988) had a trading amount of 1.878 billion with a net buy of -107 million [2] - Tencent Holdings (00700) recorded a trading amount of 1.743 billion with a net buy of 543 million [2] - China National Offshore Oil Corporation (00883) achieved a trading amount of 1.608 billion with a net buy of 142 million [2] - SMIC (00981) had a trading amount of 1.269 billion with a net buy of -266 million [2] - Xiaomi Group-W (01810) recorded a trading amount of 1.064 billion with a net buy of 295 million [2] - Yingfu Fund (02800) had a trading amount of 1.042 billion with a net buy of -1.036 billion [2] - Zijin Mining (02899) achieved a trading amount of 1.017 billion with a net buy of 528 million [2] - Laopu Gold (06181) had a trading amount of 0.928 billion with a net buy of -78.823 million [2] - JunDa Co., Ltd. (02865) recorded a trading amount of 0.812 billion with a net buy of -74.468 million [2] - China Life (02628) had a trading amount of 0.707 billion with a net buy of -107 million [2]

Group 1 - Southbound funds recorded a net sell of HKD 826 million in Hong Kong stocks for the second consecutive day [1] - Notable net purchases included Tencent Holdings at HKD 1.015 billion, Xiaomi Group at HKD 822 million, and Pop Mart at HKD 613 million [1] - Significant net sales were observed in China Mobile at HKD 1.177 billion, Zijin Mining at HKD 1.077 billion, and Alibaba at HKD 235 million [1] Group 2 - Southbound funds have consistently net bought Xiaomi for 7 days, totaling HKD 3.89621 billion [1] - Continuous net buying for China National Offshore Oil Corporation (CNOOC) for 4 days, amounting to HKD 1.09627 billion [1] - Pop Mart has seen net buying for 3 consecutive days, totaling HKD 1.73397 billion [1] - China Mobile has faced net selling for 16 days, accumulating to HKD 12.89303 billion [1]

Group 1 - The core point of the article is that JPMorgan has increased its long position in Alibaba Group Holding Limited from 8.08% to 8.15% as of January 21, 2026 [1] Group 2 - The increase in JPMorgan's holdings indicates a growing confidence in Alibaba's market position and potential future performance [1] - This change in ownership percentage reflects ongoing interest from institutional investors in Alibaba [1] - The data suggests a slight but notable shift in investment strategy by JPMorgan regarding Alibaba [1]