Investment Rating - Industry Investment Rating: Overweight [1] Core Insights - The report highlights the strong momentum in AI development both domestically and internationally, with a focus on opportunities in the Hongmeng ecosystem [4][43] - Google announced the release of the upgraded Gemini 2.5 Pro AI model, which is available through Gemini API and Google's Vertex AI and AI Studio platforms [3][45] - Alibaba Cloud has open-sourced the Qwen3 series models, which have shown superior performance in various benchmarks compared to well-known models like OpenAI's [5][44] Market Performance Review - From May 6 to May 9, 2025, the Shanghai Composite Index rose by 1.92%, while the ChiNext Index increased by 3.27%. The CSI 300 Index saw a rise of 2%, and the Hang Seng Tech Index fell by 1.22% [23][36] - The AI index increased by 2%, and the cloud computing index rose by 2.19%, indicating positive trends in these sectors [23][36] AI Developments - Google is set to launch the NotebookLM mobile app on May 20, 2025, which is currently available for pre-order [43] - Tencent has released and open-sourced a new multimodal video generation tool called Hunyuan Custom, which integrates various input modalities [44] - Kimi has launched a new general audio model, Kimi-Audio, supporting multiple audio-related tasks [44] Semiconductor Sector - TSMC reported a sales figure of 349.57 billion New Taiwan Dollars for April 2025, marking a year-on-year increase of 48.1% and a month-on-month increase of 22.2% [45]

Economic Indicators - China's CPI for April shows a year-on-year decline of 0.1%, marking the third consecutive month of deflation[12] - The PPI for April decreased by 2.7% year-on-year, continuing a 31-month trend of factory deflation[12] - China's trade balance for April was $96.18 billion, with exports increasing by 8.1% and imports decreasing by 0.2% year-on-year[17] Market Performance - The Shanghai Composite Index fell by 0.3% to 3342 points, while the Shenzhen Component dropped by 0.69%[1] - The Hang Seng Index rose by 0.4% to 22867.74 points, with the Hang Seng Tech Index declining by 0.93%[1] - The total market turnover in Hong Kong decreased to 1616.286 billion HKD[1] Trade Negotiations - The US and China reported "substantial progress" in trade negotiations held in Geneva, with a joint statement expected to be released[12] - Both parties agreed to establish a trade consultation mechanism to address economic concerns[12] Corporate Developments - CATL plans to raise approximately 30.7179 billion HKD through an IPO, with 90% of the funds allocated for projects in Hungary[10] - China Overseas Land & Investment reported a 7.5% year-on-year decline in property sales for April, totaling approximately 201.64 billion RMB[14]

【太平洋科技快讯】5月9日，腾讯混元正式推出并开源一款全新的多模态定制化视频生成工具—— Hunyuan Custom，该工具基于混元视频生成大模型(Hunyuan Video)打造。 Hunyuan Custom 的核心优势在于其强大的多模态融合能力。它能够同时处理文本、图像、音频、视频 等多种输入形式，并将其转化为连贯、自然的视频内容。相比传统视频生成模型，Hunyuan Custom 在 生成质量和控制力方面都有着显著提升。 Hunyuan Custom 具备强大的扩展能力。在音频驱动模式下，用户可以上传人物图像并配上音频语音， 模型便可生成人物在任意场景中说话、唱歌或进行其他音视频同步表演的效果，广泛适用于数字人直 播、虚拟客服、教育演示等场景。在视频驱动模式下，Hunyuan Custom 支持将图片中的人物或物体自 然地替换或插入到任意视频片段中，进行创意植入或场景扩展，轻松实现视频重构与内容增强。 此外，Hunyuan Custom 提供了多种视频生成模式，包括单主体视频生成、多主体视频生成、单主体视 频配音以及视频局部编辑等。其中，单主体生成能力已经开源并在混元官网上线，用户可以在"模型广 场 ...

Group 1 - Tencent launched and open-sourced a new multimodal customized video generation tool called Hunyuan Custom, which is based on the Hunyuan Video model and offers superior consistency effects compared to existing open-source solutions [1] - Nvidia has open-sourced its Open Code Reasoning (OCR) model suite, which includes three parameter sizes: 32B, 14B, and 7B, all released under the Apache 2.0 license and available for download on Hugging Face [2] - The 32B model is designed for high-performance inference and research scenarios, while the 14B model balances computational demands with strong reasoning capabilities, and the 7B model is suitable for resource-constrained environments [2] Group 2 - Amazon Web Services (AWS) is secretly developing an AI programming tool codenamed "Kiro," which aims to generate code in "near real-time" through a multimodal interface with an AI Agent [3] - Kiro's key features include real-time access to knowledge bases and third-party plugins, covering the entire software development process from technical design documentation to real-time code writing and vulnerability detection [3] - OpenAI has introduced the first "deep research connector" for ChatGPT, allowing developers to connect their GitHub repositories for in-depth analysis of code structures and generation of detailed research reports [4]