Alibaba Intelligence

9.29 On September 28, SuperCLUE, the comprehensive evaluation benchmark for Chinese general-purpose large models, released its September 2025 ranking. The data shows that while top-tier international models continue to maintain their lead, domestic large models are closing the gap at an astonishing pace, demonstrating strong competitiveness across multiple key capabilities.

10.22 在行业普遍倾向闭源的当下，通义千问通过持续扩充开源模型家族，正推动AI能力从可用向好用易用演进，加速多模态大模型在千行百业中的真实应用与价值释放

1.27 阿里云在毫无预热的情况下突然发布千问系列最强推理模型Qwen3-Max-Thinking，直接对标OpenAI GPT-5.2和Google Gemini 3 Pro，在多项基准测试中展现出媲美甚至超越国际顶流的性能。此举被业界视为中国AI从"跟跑"到"并跑"的关键转折点，标志着国产大模型在核心算法层面正式迈入全球第一梯队

4.29 这是国内首个“混合推理模型”，将“快思考”与“慢思考”集成进同一个模型，大大节省算力消耗。

7.31 ​On July 31, ByteDance officially released its next-generation video generation model Seedance 2.5, doubling the single-generation length from 15 to 30 seconds, supporting up to 50 multimodal reference inputs and native 4K quality. The model is rolling out to J Dream and Doubao Pro, with API access to Volcano Ark coming soon.

6.24 On June 24, at the FORCE Power Conference, ByteDance officially released the Doubao 2.1 Pro large model, with core metrics that shook the industry: programming capabilities on par with Anthropic's Claude Opus 4.7, and overall usage costs reduced by nearly 80% compared to the Claude Opus 4.6–4.8 series. At the same time, the company previewed that its video generation model Seedance 2.5 will officially launch in early July,

12.2 今日，阿里巴巴旗下的通义实验室宣布，其官方应用千问App已完成重大升级，正式向国内用户免费开放两款重量级多模态AI模型被誉为国内最强的视频生成模型通义万相Wan2.5，以及被称为全球最强开源模型的Qwen-Image的独家满血特供版

4.1 该模型采用混合注意力MoE架构，支持113种语言方言识别256K长上下文，具备VibeCoding能力，可通过自然语言生成复杂代码，目前已在阿里云百炼上线，每百万Tokens输入成本不到0.8元

6.11 在HuggingFace最新发布的开放权重模型贡献榜单中，来自中国的Qwen和DeepSeek团队双双跻身前15名，展示了中国在开源AI领域的技术实力与全球影响力

12.20 Based on multiple dimensions—including general model capabilities, domain-specific capabilities, vendor execution capabilities, solution richness, strategy and innovation, and market influence—Omdia conducted a comprehensive assessment of the top 12 leading Chinese model vendors, categorizing them by score into three quadrants: Leaders, Challengers, and Aspirants.

10.28 MiniMax-M2 在-BenchSWE-Bench和BrowseComp中的性能表明，对于依赖能够规划执行和验证复杂工作流程的AI系统的组织来说，这些系统具有特殊的优势，这些系统是企业环境中代理和开发人员工具的关键功能

5.10 据TheInformation最新报道，DeepSeek正寻求完成一轮高达500亿元人民币约73.5亿美元的融资，若这把成了，这不仅将刷新中国AI公司单轮融资的最高纪录，更会将这家成立仅三年的公司，一举推上3500亿元人民币约515亿美元的估值高位

6.9 维度阿里巴巴字节跳动核心策略Token工业化+企业级服务全栈自研+消费者生态组织架构TokenFoundryCEO直管火山引擎+Seed团队技术路线通义千问+开源生态Seedance+Seedream+豆包商业场景电商金融物流短视频内容创作广告资本投入未公开具体数字700亿美元5年

8.28 On August 27, TIME magazine officially released the 2026 TIME100 AI list, its fourth annual ranking of the 100 most influential people in artificial intelligence. Seventeen representatives graced the magazine covers, including OpenAI CEO Sam Altman, xAI founder Elon Musk, Alibaba Group CEO Wu Yongming, and Stanford Professor Fei-Fei Li.

6.30 该模型不仅能精准解析文本图像音频等不同模态的数据，还能实现跨模态的交互与创作，例如根据文字描述生成高质量图像，或对现有图像进行智能编辑优化

9.1 On August 31, the Intelligent Chemical Industry Large Model 3.0 Pro was officially released, co-developed by the Dalian Institute of Chemical Physics (DICP) of the Chinese Academy of Sciences, iFLYTEK, Alibaba Cloud, and other partners. As the latest iteration of China's first large model for the chemical industry, 3.0 Pro marks a significant leap from a "professional knowledge Q&A model" to an "intelligent task execution system for chemical engineering," upgrading from a "professional assis

5.1 本周，谷歌与通用汽车接连发布重磅消息：Google Gemini AI助手将正式进入搭载Google built-in系统的汽车，覆盖通用汽车旗下约400万辆2022款及更新的凯迪拉克、雪佛兰、别克和GMC车型 。这不仅是谷歌在车载AI领域最大规模的一次部署，更标志着智能座舱从“功能时代”向“生成式AI时代”的范式转移

5.11 沙特阿联酋等国家，已经通过主权财富基金累计投资超过3200亿美元布局AI数据中心超级算力与能源基础设施，希望借助能源+资本优势进入全球AI核心竞争

8.21 阿里云8 月 21 日宣布，阿里云、超星尔雅协同北京大学、南京大学、复旦大学、上海交通大学、浙江大学五所高校名师，共同推出的《 AI 通识公益系列课程》“动手学 AI：人工智能通识与实践”，将于 9 月 1 日面向全国所有高校、所有专业的师生正式开放

8.27 面向不同场景，Qoder提供编程和通用两种模式编程模式适合代码相关任务，通用模式让非开发者也能借助Coding和工具能力完成任务