本文源自：市场资讯 作者：情报员 声明：市场有风险，投资需谨慎。本文为AI基于第三方数据生成，仅供参考，不构成个人投资建议。 国家知识产权局信息显示，腾讯科技（深圳）有限公司申请一项名为"业务系统的测试方法、装置、电 子设备及存储介质"的专利，公开号CN121029577A，申请日期为2024年5月。专利摘要显示，本申请涉 及计算机领域，公开了一种业务系统的测试方法，应用于第一应用的第一应用服务端，方法包括：获取 目标接口的测试配置信息，测试配置信息包括目标接口的生效参数和重放参数；若目标接口的生效参数 指示目标接口在第一业务系统中可被调用，从候选请求集合中获取请求目标接口的目标请求；按照目标 接口的重放参数，向第一业务系统中的目标接口和第二业务系统中的目标接口重放目标请求；获取第一 业务系统响应于目标请求返回的第一响应数据和第二业务系统响应于目标请求返回的第二响应数据；根 据目标接口对应的第一响应数据和对应的第二响应数据，确定第一业务系统的测试结果，实现了以接口 为单位的测试。 天眼查资料显示，腾讯科技（深圳）有限公司，成立于2000年，位于深圳市，是一家以从事软件和信息 技术服务业为主的企业。企业注册资本 ...

深圳 3D 打印赛道的 "内卷"，从资本布局的密集程度可见一斑。作为互联网巨头的腾讯，早已在该领域 埋下 "双保险"—— 公开信息显示，腾讯不仅是创想三维的股东，还曾被传参投拓竹科技新一轮融资 （估值或达 100 亿美元），尽管拓竹创始人陶冶后续回应 "目前没进行中的融资"，但双方此前的合作 已十分紧密：2025 年 7 月，拓竹旗下 3D 模型平台 MakerWorld 全面接入腾讯混元 3D 生成模型，借助 生成式 AI 降低用户建模门槛，为消费级 3D 打印机的市场爆发铺路。 另一边，大疆的入局则带着更强的 "技术协同" 属性。2025 年 11 月，天眼查数据显示大疆持有智能派 5% 股权，大疆相关人士对媒体表示，"投资基于看好消费级 3D 打印技术发展潜力，符合公司对创新科 技的前瞻性布局"。而智能派联合创始人陈波在接受新闻采访时也直言，相较于单纯财务投资，更看重 大疆的技术能力补足，"融资后将对标大疆 All In 产品研发，补齐生态建设"。 2025 年 8 月，深圳 3D 打印企业创想三维正式向港交所递交主板上市申请，冲击 "港股消费级 3D 打印 第一股"；仅 1 个多月后，同为深圳本土企业 ...

Core Insights - The report by Goldman Sachs analyzes the competitive landscape of China's AI industry, focusing on the strategic choices of major players like Alibaba, ByteDance, and Tencent [2][6][18]. Group 1: Alibaba's Strategy - Alibaba is pursuing a "full-stack" approach similar to Google's, with a significant capital expenditure increase of 80% year-on-year, reaching RMB 32 billion [6][7]. - The company aims to build a robust AI infrastructure through vertical integration of "base models + multimodal capabilities," despite challenges in chip supply [6][7]. - Alibaba Cloud's external revenue grew by 29% year-on-year in the September quarter, with AI-related revenue achieving triple-digit growth for nine consecutive quarters [7][8]. Group 2: ByteDance's Approach - ByteDance is leveraging its dominance in consumer applications to enhance its foundational infrastructure, with daily token usage surpassing 30 trillion, approaching Google's 43 trillion [10][14]. - The company's education app Gauth has seen a 394% year-on-year increase in monthly revenue, indicating strong market performance [11]. - ByteDance's Volcano Engine holds a 49.2% market share in the public cloud market for large models, showcasing its competitive edge [14]. Group 3: Tencent's Position - Tencent has adopted a more restrained approach, reducing capital expenditures while focusing on integrating AI capabilities into its extensive social and payment ecosystem [15][17]. - The company has integrated its AI assistant "Yuanbao" into WeChat Pay, enhancing operational efficiency for small and medium-sized businesses [17]. Group 4: US-China AI Competition - The competition between the US and China in AI has entered a "dynamic alternation" phase, with Chinese models expected to rapidly iterate and catch up within 3-6 months following significant advancements in US models [4][19]. - Chinese companies are noted for their resilience and speed, with many leveraging open-source models to enhance their capabilities [19]. Group 5: Valuation Insights - Goldman Sachs indicates that the current state of the Chinese AI sector does not reflect a bubble, with expected P/E ratios for Tencent and Alibaba at 21x and 23x respectively, lower than those of major US tech companies [20].

以往跨境游客购买手信时面临"大包小包"、"携带不便"等痛点，而本次活动的线下门店——"澳门手信微信礼物体验店"汇聚了超20家澳门商家的特 色手信，并创新引入微信礼物功能，游客在店内只需扫描商品"送礼码"下单，即可借助微信蓝包在线上秒送澳门手信给亲朋好友，免去随身携带入境的 麻烦。 港澳地区一直是内地游客出境游首选目的地，今年1月至10月澳门入境旅客超3000万人次，其中超七成为内地游客。为更好服务内地游客赴澳消 费，推动中小商家利用微信生态创新增长，11月28日在澳门经济及科技发展局的支持下，澳门直播协会联合腾讯正式启动微信小店"礼遇澳门"直播电商 好物节，并在澳门最繁华的商业街区——议事厅前地广场附近澳门何老桂巷5号落地首个微信礼物线下体验店，超20家澳门商家的特色手信汇聚线下体 验店，让跨境游客体验"一站式""零负担"买送澳门手信。 中央驻澳门联络办公室、澳门经济及科技发展局、澳门直播协会、腾讯公司等代表为澳门手信微信礼物体验店揭牌 （澳门手信微信礼物线下体验店正式开业） 澳门微信礼物体验店采用在澳门下单、内地发货的"前店后仓"模式，是数字经济和实体经济深度融合发展的生动实践。这一模式可以为商家提升运 营 ...

Core Insights - The report by Goldman Sachs highlights the intense competition in China's AI sector, focusing on the strategic choices of major players like Alibaba, ByteDance, and Tencent, and suggests a new normal of "dynamic alternation" in the US-China AI competition [1][2] Group 1: Alibaba's Strategy - Alibaba is adopting a "full-stack" approach similar to Google's, with a significant increase in capital expenditure, which surged by 80% year-on-year to reach 32 billion RMB in the September quarter [3][4] - The company's cloud revenue grew by 29% year-on-year, with AI-related revenue achieving triple-digit growth for the ninth consecutive quarter, and is expected to accelerate to 38% growth in the December quarter [4][6] Group 2: ByteDance's Approach - ByteDance is leveraging its massive traffic advantage, with a daily token consumption of 30 trillion, approaching Google's 43 trillion, and significantly surpassing competitors like Baidu [9][13] - The company's application "Doubao" leads in domestic AI application activity, while its overseas education app Gauth saw a 394% year-on-year increase in monthly revenue [9][13] Group 3: Tencent's Strategy - Tencent is maintaining a conservative approach, reducing capital expenditure while focusing on seamlessly integrating AI capabilities into its extensive social and payment ecosystem [14][15] - The company has integrated its AI assistant "Yuanbao" into WeChat Pay, enhancing operational efficiency for small and medium-sized businesses [15] Group 4: US-China AI Competition - The report outlines a "dynamic catch-up" cycle in the US-China AI competition, where Chinese models typically follow significant advancements in US models within 3-6 months [16][17] - Chinese companies are noted for their resilience and aggressive cost control, with many leveraging open-source models to enhance their capabilities [17] Group 5: Valuation Insights - Goldman Sachs indicates that the current state of the Chinese AI sector does not reflect a bubble, with projected P/E ratios for Tencent and Alibaba at 21x and 23x respectively, lower than those of major US tech companies [18]

本月获批的国产游戏版号里，值得注意的有腾讯游戏《无境撤离》、B站《闪耀吧！噜咪》、完美世界 《梦幻新诛仙：轻享》、西山居投资的《星砂岛》、雷霆游戏《再世仙途》、四三九九《造梦西游之黎 尤浩劫篇》、恺英网络《冰雪王者》、冰川网络《小小护卫队》、胖布丁《绿水镇》等。 11月游戏版号发放，版号总量已超1600款 11月27日，国家新闻出版署公布了2025年11月国产网络游戏及进口网络游戏审批信息，共有184款游戏 过审。国产游戏版号178个，其中客户端游戏2款，移动、客户端双端游戏7款，客户端、游戏机 （PS5）游戏1款，移动、客户端、网页三端游戏1款，其余均为移动游戏；进口游戏版号6个，其中客 户端游戏1款，移动、客户端双端游戏1款，其余均为移动游戏。此外，还有7款游戏做出了信息变更。 截至11月，2025年共计下发游戏版号1624个，其中国产游戏版号1532个，进口游戏版号92个。近1年 来，国产游戏版号每月过审数量均在110款以上，今年下发的版号数量相比去年有明显增加，仍然呈现 稳定增长趋势。 字节跳动或计划出售沐瞳，买家为沙特集团 11月26日，据彭博社报道，字节跳动正与沙特Savvy Games Grou ...

Core Points - A significant fire occurred in Hong Kong's Tai Po district, resulting in major casualties and prompting over 100 companies and foundations to donate for emergency relief and community recovery efforts, with total donations exceeding HKD 1.2 billion [2][4]. Donation Summary - Major companies such as Yuexiu Group donated HKD 10 million, while China Overseas and China State Construction contributed HKD 20 million [3]. - Tencent donated HKD 30 million, and Alibaba made an initial donation of HKD 20 million [3]. - Other notable contributions include HKD 30 million from the Li Ka Shing Foundation and HKD 12 million from the Chaozhou Association [4]. - The total amount of donations has surpassed HKD 1.2 billion as of the latest reports [4].

此外，在金融细分领域当中，零售银行、企业金融、证券、保险等也对大模型有差异化需求。 胡利明表示，不同细分领域解决方案的进步是"螺旋式迭代"的。以腾讯云为例，首先依托其基础大模 型"混元"的迭代提升能力，然后，针对银行、证券、保险等不同领域，如证券领域投研、投顾，客服等 细分场景，以不同的细节要求，基于自研的 TI 平台、训练专属领域小模型。通过 "小模型 + 大模型" 的 组合，搭配工程化开发与串接，实现智能体的开发及部署上线。基础大模型和场景应用工程的迭代进 步，都会推动AI 应用的效果提升。 专题：2025年大湾区交易所科技大会 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！ 在深交所、港交所、广期所共同举办的"2025年大湾区交易所科技大会"中，腾讯云副总裁胡利明在媒体 采访环节对新浪财经表示，在AI大模型加速渗透各行业的背景下，金融行业作为数据密集、风险敏感 的领域，其对大模型的要求呈现高门槛特征。 胡利明表示，金融行业对大模型提出了更为严苛的要求，主要集中在合规性与内容准确性两个方面。金 融行业的合规监管体系极为严格，不同业务领域、不同牌照持有方均需遵循相关的法律法规与行业 ...

11月26日下午，香港大埔宏福苑发生五级火灾，灾情严峻，牵动各方关注。腾讯、阿里、京东、滴滴、 字节跳动、拼多多、小米等互联网大厂捐款捐物，驰援香港。 微信小游戏WeCare公益计划也同步上线了"守望相助，驰援香江"救灾捐赠项目。小游戏开发者可以在 微信公众号后台选择该项目，善款将从游戏收益中快速捐赠至基金会爱心账户，并获得捐赠证明和相关 收据。截至11月29日上午9点，微信小游戏开发者已经捐赠善款超过230万元。 此外，腾讯旗下WeChat港币钱包紧急上线支援大埔火灾救援捐款专区，开设特区政府"大埔宏福苑援助 基金"数转快识别码捐赠及银行账户捐赠直达入口，并涵盖仁爱堂、仁济医院、香港公益金、香港圣公 会福利协会、东华三院等9家慈善机构捐款专项，还设置支付后捐款引导提示，方便香港市民捐款支 援，过程中不收取任何行政费用或手续费，善款全数转交相关机构。 同时，微信支付也上线了捐赠入口，用户可以在支付完成后，直接进入腾讯公益平台的项目捐赠页面， 为救援项目献一份爱心。 （文章来源：南方都市报） 11月27日上午，腾讯公益慈善基金会（香港）启动首批捐款1000万港元，用于受灾居民的紧急救援、过 渡安置、生活物资补给 ...

Core Insights - Tencent's HunyuanOCR model is a commercial-grade, open-source, lightweight OCR-specific visual language model with 1 billion parameters, combining native ViT and lightweight LLM architectures [1] - The model excels in perception capabilities (text detection and recognition, complex document parsing) and semantic abilities (information extraction, text-image translation), winning the ICDAR 2025 DIMT challenge and achieving SOTA results on OCRBench for models under 3 billion parameters [2] Model Performance and Popularity - HunyuanOCR ranks in the top four on Hugging Face's trending list, has over 700 stars on GitHub, and was integrated by the vllm official team on Day 0 [3] Team Achievements - The HunyuanOCR team has achieved three major breakthroughs: 1. Unified efficiency, supporting various tasks like text detection, complex document parsing, and visual question answering within a lightweight framework [5] 2. Simplified end-to-end architecture, eliminating dependencies on pre-processing and reducing deployment complexity [6] 3. Data-driven innovations using high-quality data and reinforcement learning to enhance OCR task performance [8] Core Technology - HunyuanOCR focuses on lightweight model structure design, high-quality pre-training data production, application-oriented pre-training strategies, and task-specific reinforcement learning [11] Lightweight Model Structure - The model employs an end-to-end training and inference paradigm, requiring only a single inference to achieve complete results, avoiding common issues of error accumulation in traditional architectures [14][19] High-Quality Data Production - The team built a large-scale multimodal training corpus with over 200 million "image-text pairs," covering nine core real-world scenarios and over 130 languages [21] Pre-Training Strategy - HunyuanOCR uses a four-stage pre-training strategy focusing on visual-language alignment and understanding, with specific stages dedicated to long document processing and application-oriented training [29][32] Reinforcement Learning Approach - The model innovatively applies reinforcement learning to enhance performance, using a hybrid strategy for structured tasks and LLM-based rewards for open-ended tasks [36] Data Quality and Reward Design - The data construction process emphasizes quality, diversity, and difficulty balance, utilizing LLM to filter low-quality data and ensuring effective training [39] - Adaptive reward designs are implemented for various tasks, ensuring precise and verifiable outputs [40][42]