![[7afd2f54-0711-4894-b34d-8d606a558326.mp3]]

中国大模型数量领跑全球，总数已超1500个

简报：

• 根据世界人工智能大会最新数据，全球已发布3755个大模型，中国企业贡献了1509个，数量位居全球首位。
• 我国已成为全球人工智能专利最大拥有国，占比达60%，预计到2024年，人工智能产业规模将突破7000亿元。
• 近期，阿里巴巴、商汤科技、京东、蚂蚁数科等公司均在大会上发布了各自的新一代大模型产品，如“通义千问”、“日日新V6.5”、“JoyAI”和“Agentar-Fin-R1”。
• 截至2025年6月，我国使用生成式人工智能产品回答问题的用户比例高达80.9%。

相关链接：

• 我国大模型数量突破1500，全球领先地位凸显
• 商汤科技重磅发布：日日新6.5大模型与 “悟能” 智能平台引领AI新潮流！
• 阿里巴巴重磅发布三款开源大模型，未来AI将更加智能！
• 蚂蚁数科联合多家机构发布Finova大模型金融应用评测基准
• 京东宣布其大模型品牌正式升级为JoyAI 并发布全新智能品牌JoyInside

China Leads Globally in Large Language Models, Total Exceeds 1500

Briefing:

• According to the latest data from the World Artificial Intelligence Conference (WAIC), a total of 3755 large models have been released globally, with Chinese enterprises contributing 1509, ranking first worldwide. 🚀
• China has become the world's largest holder of AI patents, accounting for 60% of the global total. By 2024, the AI industry scale is projected to exceed 700 billion RMB.
• Recently, companies like Alibaba, SenseTime, JD.com, and Ant Group Digital Technologies have unveiled their new generation of large model products at the conference, including 'Tongyi Qianwen,' 'SenseNova V6.5,' 'JoyAI,' and 'Agentar-Fin-R1.' ✨
• As of June 2025, the proportion of users in China utilizing generative AI products to answer questions has reached an impressive 80.9%. 📈

Related Links:

• 我国大模型数量突破1500，全球领先地位凸显
• 商汤科技重磅发布：日日新6.5大模型与 “悟能” 智能平台引领AI新潮流！
• 阿里巴巴重磅发布三款开源大模型，未来AI将更加智能！
• 蚂蚁数科联合多家机构发布Finova大模型金融应用评测基准
• 京东宣布其大模型品牌正式升级为JoyAI 并发布全新智能品牌JoyInside

特斯拉机器人明确入华时间表：2025年进C端，2026年量产

简报：

• 特斯拉相关人士透露，第三代机器人已在美国工厂实测，预计2025年面向中国C端市场推出，进入家庭等消费场景。
• 该机器人预计于2026年实现量产，当年计划生产10万台，并计划5年内实现100万台的年产能。
• 马斯克在财报会议中表示，即将量产的机器人将采用全新的V3设计，该设计已无需根本性改变，并将在协调性和执行叠衣服等复杂家庭任务的能力上实现突破。

相关链接：

• 消息称特斯拉第三代机器人即将进入中国 C 端市场，预计明年实现量产
• 国家地方共建人形机器人创新中心签约首批开放基金及分训练场
• 茅台旗下基金、中信证券投资等入股乐聚机器人公司
• 国产机器人走向全球，智元拟进军海外市场

Tesla Bot's Clear China Entry Timeline: 2025 for Consumer Market, 2026 for Mass Production 🤖

Brief:

• Tesla insiders reveal that the third-generation robot has been tested in US factories and is expected to be launched for China's C-end (consumer) market in 2025, entering household and other consumption scenarios.
• The robot is expected to achieve mass production in 2026, with a planned output of 100,000 units that year, and an annual production capacity target of 1 million units within 5 years. 🚀
• Elon Musk stated in an earnings call that the soon-to-be mass-produced robot will feature an all-new V3 design, which requires no fundamental changes and will achieve breakthroughs in coordination and the ability to perform complex household tasks like folding clothes. ✨

Related Links:

• Report: Tesla's Third-Generation Robot Set to Enter China's C-End Market, Mass Production Expected Next Year
• National and Local Governments Sign Agreements for First Batch of Open Funds and Sub-training Fields for Humanoid Robot Innovation Center
• Kweichow Moutai's Fund, CITIC Securities Investment, and Others Invest in Leju Robot Company
• Chinese Robots Go Global: Zhiyuan Plans to Enter Overseas Markets

腾讯开源业界首个可交互3D世界生成模型

简报：

• 腾讯于2025年7月27日在世界人工智能大会上正式发布混元3D世界模型1.0，并宣布全面开源。
• 该模型是业界首个支持沉浸式漫游、交互与仿真的开源世界生成模型，可根据文字或图片输入，在数分钟内生成可漫游的3D场景。
• 模型采用创新的“语意层次化3D场景表征及生成算法”，能够输出兼容Unity、Unreal Engine等主流工具的3D Mesh资产，便于开发者进行二次编辑。
• 该模型支持高精度的360°全景场景生成和可交互的漫游场景功能，后者具备智能范围判断以防止“穿墙”。
• 腾讯为用户提供了每日20次的免费生成额度，并已在GitHub、Hugging Face等平台开源模型及代码。

相关链接：

• 项目主页
• 体验地址
• Hugging Face 模型地址
• Github 项目地址

Tencent Open-Sources Industry's First Interactive 3D World Generation Model

Briefing:

• Tencent officially launched its Hunyuan 3D World Model 1.0 at the World Artificial Intelligence Conference on July 27, 2025, announcing its full open-sourcing.
• This model is the industry's first open-source world generation model supporting immersive roaming, interaction, and simulation. It can generate navigable 3D scenes within minutes based on text or image inputs. 🚀
• The model employs an innovative "Semantic Hierarchical 3D Scene Representation and Generation Algorithm," capable of outputting 3D Mesh assets compatible with mainstream tools like Unity and Unreal Engine, facilitating secondary editing for developers.
• It supports high-precision 360° panoramic scene generation and interactive roaming capabilities, with the latter featuring intelligent boundary detection to prevent "clipping" or "passing through walls." 🏞️
• Tencent offers users a daily free generation quota of 20 times and has open-sourced the model and code on platforms such as GitHub and Hugging Face. ✨

Related Links:

• Project Homepage
• Experience Address
• Hugging Face Model Address
• Github Project Address

阿里开源AI智能体WebSailor，专攻复杂信息检索

简报：

• 阿里巴巴通义实验室开源了一款名为WebSailor的AI智能体框架，旨在应对信息检索挑战。
• WebSailor的核心技术结合了复杂任务生成和强化学习，通过两阶段训练方法优化模型。
• 在BrowseComp等基准测试中，WebSailor的性能表现已超越所有现有开源智能体，并能与一些闭源模型相媲美。
• 该项目在GitHub上已获得超5000颗星，是每日增长率最高的项目之一。

相关链接：

• WebSailor开源地址

Alibaba Open-Sources AI Agent WebSailor, Specializing in Complex Information Retrieval

Briefing:

• Alibaba Tongyi Lab has open-sourced an AI agent framework named WebSailor, designed to address complex information retrieval challenges.
• WebSailor's core technology combines complex task generation with reinforcement learning, optimizing the model through a two-stage training approach.
• In benchmarks like BrowseComp, WebSailor's performance has surpassed all existing open-source agents and can rival some closed-source models. 🚀
• The project has already garnered over 5,000 stars on GitHub, making it one of the fastest-growing projects daily. ✨

Related Link:

• WebSailor GitHub Repository 🔗

以色列研发新型可穿戴设备，能实时监测血糖和药物浓度

简报：

• 以色列理工学院及同行联合开发出一款新型可穿戴设备，能够实时监测糖尿病患者体内的血糖水平和药物浓度。
• 该技术未来也可通过调整传感机制，推广应用于其他对药物剂量要求严格的慢性疾病治疗管理。

相关链接：

• 新型可穿戴设备能实时监测体内血糖和药物浓度

Israel Develops New Wearable Device for Real-Time Blood Glucose and Drug Concentration Monitoring

Brief:

• Researchers from the Technion – Israel Institute of Technology and colleagues have co-developed a new wearable device capable of real-time monitoring of blood glucose levels and drug concentrations in diabetic patients. 💡
• In the future, this technology could also be adapted to other chronic diseases requiring strict drug dosage management by adjusting the sensing mechanism.

Related Links:

• 新型可穿戴设备能实时监测体内血糖和药物浓度