![[378c7beb-8df5-45ac-ad0a-e5d34ba5b10e.mp3]]

鸿海精密申请注册“FoxBrain”商标，聚焦AI大模型领域

简报：

• 近日，富士康母公司鸿海精密工业股份有限公司申请注册“FoxBrain”商标，国际分类为科学仪器，目前处于注册申请中。
• “FoxBrain”是鸿海集团旗下鸿海研究院自主研发的首款繁体中文AI推理大模型，于2024年3月发布，基于Meta Llama3.1架构开发优化，专注于数学推演、逻辑分析与代码生成领域。

相关链接：

• 鸿海精密申请注册“FoxBrain”商标

Foxconn Technology Group Applies to Register "FoxBrain" Trademark, Focusing on AI Large Models 🧠

Briefing:

• Recently, Foxconn Technology Group (Hon Hai Precision Industry Co., Ltd.), the parent company of Foxconn, applied to register the "FoxBrain" trademark. Its international classification is "scientific instruments," and the application is currently under review.
• "FoxBrain" is the first traditional Chinese AI inference large model independently developed by Hon Hai Research Institute, a subsidiary of the Hon Hai Group. It was released in March 2024, developed and optimized based on the Meta Llama 3.1 architecture, focusing on mathematical reasoning, logical analysis, and code generation. 🤖

Related Links:

• Hon Hai Precision Industry Applies to Register "FoxBrain" Trademark 💡

OpenAI推出高价企业级AI定制服务，与咨询巨头直接竞争

简报：

• OpenAI近日正式推出面向大型企业的高端AI咨询服务，每位客户收费标准至少为1000万美元。
• 该服务涵盖模型（如GPT-4o）的定制、数据标注等，工程师与企业深度合作开发专属应用，例如企业专用聊天机器人。
• 数据标注工作部分拟外包，以提升效率和准确性，合作方包括Snorkel AI和Surge AI等专业公司。
• OpenAI现有客户包括美国国防部和东南亚科技企业Grab，目标助力企业实现数字化转型。
• 此举标志OpenAI与Palantir、Accenture等行业巨头直接竞争，抢占企业级AI解决方案市场。

相关链接：

• OpenAI启动高端企业AI咨询服务，客户收费至少1000万美元
• OpenAI企业AI咨询业务将与巨头竞争

OpenAI Rolls Out High-Priced Enterprise AI Customization Service, Directly Challenging Consulting Giants

Brief:

• OpenAI has officially launched a premium AI consulting service for large enterprises, with a minimum charge of $10 million per client. 💰
• The service covers model customization (e.g., GPT-4o) and data labeling, with engineers working closely with enterprises to develop exclusive applications, such as company-specific chatbots. 🤖
• Parts of the data labeling work are planned to be outsourced to enhance efficiency and accuracy, with partners including specialized companies like Snorkel AI and Surge AI.
• OpenAI's existing clients include the U.S. Department of Defense and Southeast Asian tech company Grab, aiming to help enterprises achieve digital transformation.
• This move marks OpenAI's direct competition with industry giants like Palantir and Accenture, vying for a share of the enterprise AI solutions market. 📈

Related Links:

• OpenAI Launches High-End Enterprise AI Consulting Service, Charging Clients at Least $10 Million
• OpenAI's Enterprise AI Consulting Business to Compete with Giants

智谱AI获浦东10亿战略投资并开源视觉语言大模型GLM-4.1V-Thinking

简报：

• 2025年7月2日，在上海浦东张江科学会堂举办的智谱开放平台产业生态大会上，浦东创投集团和张江集团宣布联合向智谱AI战略投资总额10亿元，首笔交割已完成，三方将共建人工智能新型基础设施。
• 智谱同期开源发布新一代通用视觉语言模型GLM-4.1V-Thinking，该模型支持多模态输入（如图像、视频、文档），在28项权威评测中23项取得10B级最佳成绩，18项指标超越72B量级模型，提升了通用人工智能（AGI）的发展进程。
• 该模型现已于Hugging Face与魔搭社区同步开源，并应用于图文理解、数学科学推理、视频理解等多个领域。

相关链接：

• IT之家相关报道
• aibase.com新闻稿

Zhipu AI Secures 1 Billion Yuan Strategic Investment from Pudong, Open-Sources Vision-Language Model GLM-4.1V-Thinking

Briefing:

• On July 2, 2025, at the Zhipu Open Platform Industry Ecosystem Conference held at Zhangjiang Science Hall in Pudong, Shanghai, Pudong Venture Capital Group and Zhangjiang Group jointly announced a strategic investment of 1 billion RMB in Zhipu AI. The first installment has been completed, and the three parties will jointly build new AI infrastructure. 💰
• Concurrently, Zhipu open-sourced and released its new generation general vision-language model, GLM-4.1V-Thinking. This model supports multimodal inputs (such as images, videos, documents), achieved the best 10B-level results in 23 out of 28 authoritative evaluations, surpassed 72B-scale models in 18 metrics, and advanced the development process of Artificial General Intelligence (AGI). ✨
• The model is now open-sourced simultaneously on Hugging Face and ModelScope communities and applied in various fields, including image-text understanding, mathematical-scientific reasoning, and video understanding. 🚀

Related Links:

• IT Home Report
• aibase.com Press Release

Capital One智能代理大幅提升汽车经销商销售效率

简报：

• 在VB Transform大会上，Capital One展示了通过AI智能代理平台革新汽车销售模式，重点在于打造类人沟通与协作系统，为汽车经销商提供业务支持。
• 其AI智能代理可全天候为终端客户推荐合适车辆及贷款方案，帮助客户查看可试驾库存，有效改进客户交流与问题解决流程。
• 公司数据表明，经销商关键指标如客户参与度和销售线索提升了55%，显著提高了通过AI平台带来的业务成效。
• Capital One采用模型蒸馏等AI优化技术，并以团队专家协作的方式保障系统安全和高效，还计划将该智能代理方案拓展至旅行业务应用。

相关链接：

• Capital One借助AI革新汽车销售模式
• Capital One智能代理提升经销商客户关键指标

Capital One's AI Agent Significantly Boosts Car Dealership Sales Efficiency 🤖

Brief:

• At the VB Transform conference, Capital One showcased its AI smart agent platform, revolutionizing the car sales model. The focus is on creating human-like communication and collaboration systems to provide business support for car dealerships.
• Its AI smart agent can recommend suitable vehicles and loan options to end-customers 24/7, help customers check test-driveable inventory, and effectively improve customer communication and problem-solving processes.
• Company data indicates that key dealer metrics, such as customer engagement and sales leads, have increased by 55%, significantly boosting business results achieved through the AI platform. 📈
• Capital One employs AI optimization techniques like model distillation and ensures system security and efficiency through expert team collaboration. The company also plans to expand this smart agent solution to travel business applications. ✈️

Related Links:

• Capital One Revolutionizes Car Sales with AI
• Capital One's Smart Agent Boosts Key Dealer Customer Metrics

xAI获百亿美元资金支持，加速AI模型迭代和基础设施建设

简报：

• 马斯克旗下xAI公司近日通过50亿美元债务和50亿美元股权，总计筹集100亿美元资金，用于增强AI技术实力并与OpenAI等对手竞争。
• 这些资金将用于扩展基础设施、开发Grok旗舰AI及其最新的Grok4和Grok4Code模型，其中Grok4专注于自然语言处理和推理能力，Grok4Code主要面向开发者的代码生成。
• xAI计划建设世界最大的数据中心，并在美国田纳西州孟菲斯投入20万块GPU用于AI训练，未来计划将GPU扩展至100万个。
• 新一代Grok4模型支持大容量上下文窗口、实时多模态处理，并已开放API接口，推动其在多行业落地应用。

相关链接：

• 马斯克 xAI 公司补充 100 亿美元“弹药”，火力聚焦打造地表最强 AI
• xAI控制台新增Grok4及Grok4Code引用，标志着下一代AI模型即将发布

xAI Secures $10 Billion in Funding, Accelerating AI Model Iteration and Infrastructure Development

Brief:

• Elon Musk's xAI company recently raised a total of $10 billion ($5 billion in debt and $5 billion in equity) to strengthen its AI technological capabilities and compete with rivals like OpenAI. 🚀
• These funds will be allocated towards expanding infrastructure and developing its flagship Grok AI, including the latest Grok4 and Grok4Code models. Grok4 specializes in natural language processing and reasoning, while Grok4Code is primarily designed for developers' code generation needs.
• xAI plans to build the world's largest data center, deploying 200,000 GPUs for AI training in Memphis, Tennessee, USA, with future plans to expand to 1 million GPUs. 🌐
• The new generation Grok4 model supports large context windows and real-time multimodal processing. Its API interfaces have been opened, facilitating its application across various industries. ✨

Related Links:

• Elon Musk's xAI Company Adds $10 Billion in "Ammunition", Focusing Firepower on Building the World's Strongest AI
• xAI Console Adds Grok4 and Grok4Code References, Signaling Upcoming Release of Next-Gen AI Models