AI 资讯 · 2026-07-24

7 月 23 日，腾讯宣布混元多模态模型部门与大语言模型部门合并，成立基础模型部，统一由腾讯首席 AI 科学家姚顺雨管理，以进一步提升模型研发和协同效率，探索全模态模型的智能上限。以大语言模型为基础，腾讯混元构建了图像、视频、语音、3D 生成模型等完整的模型矩阵。2026 年以来，腾讯混元模型迭代速度不断加快，与腾讯公司业务的协同也持续增强。7 月正式上线的大语言模型 Hy3，展现出显著强于同尺寸模型且比肩参数规模 2-5 倍的旗舰模型的智能水平，已在 WorkBuddy/CodeBuddy、元宝、Marvis、ima 等多个业务接入，API 已在腾讯云 TokenHub 以及多个海外平台上线。发布一周后， Hy3 总调用量较上一代模型（Hy2）增长超 68 倍，WorkBuddy 自选模型的用户中，选择 Hy3 的占比达到 60%。混元多模态模型也在持续进化。4 月发布并开源的混元 3D 世界模型 2.0（Hy World 2.0）能够根据文字、图片、视频等不同类型输入，自动生成、重建和模拟 3D 世界，同时支持多格式 3D 资产导出，能与现有的游戏工作流无缝对接，快速生成游戏地图和

ChatGPT Voice on desktop can work with both ChatGPT Work and Codex to complete tasks and control agents.

2026年7月23日，在第十届全球跨境电商峰会暨2026全球化新品牌AI竞争力大会上，领先的AI全球营销科技平台飞书深诺Meetsocial，正式发布新一代企业级出海营销AgentOS——Marvy 2.0。Marvy 2.0不是又一款AI单点工具，而是一个会思考、会协同、会进化的营销大脑，它定位为“全球AI营销新基建”，致力于让广告增长从“可能性”走向“确定性”。大会现场，飞书深诺Meetsocial也与BytePlus正式达成官方合作伙伴关系，双方完成合作授牌仪式。Marvy 2.0将全面接入BytePlus大模型体系与Seedance音视频生成能力，获得超强技术底座加持。AI渗透率达78.5%，但出海企业更需要系统级的增长确定性文化鸿沟、合规挑战、规则政策频出、割裂的数据与后台……当下的出海企业无比渴求一把万能钥匙。AI日拱一卒的今天，市场的狂热看似为出海企业找到了新的解法，但实际产出不达预期。大会现场揭示了一个扎心的行业现实。IDC发布的《出海营销AI智能体成熟度与服务商能力白皮书》显示，出海企业营销场景的AI应用率已达78.5%，但结构性失衡严重——内容创意与本地化的AI渗透

7月24日，在 AnySearch 北京开发者沙龙活动现场，AnySearch 创始人兼CEO 薛冠群正式宣布推出“AnySearch 学生与开发者成长计划”（AnySearch Student & Developer Program）。该计划面向全球学生与开发者开放，旨在通过提供开放的 AI 搜索基础设施，支持更多 AI 创造者开展学术研究、开源创新和 AI 应用开发。 图片1 随着 AI 加速演进，AI 应用正在从简单信息交互迈向复杂任务执行，AI Agent 正在成为下一代应用的主流形态。在这一趋势下，实时、高质量的信息获取能力是 AI Agent 看见世界、完成任务的关键基础。然而，对大量学生和早期开发者而言，获取稳定、可靠的基础设施服务仍存在一定门槛。 AnySearch 希望通过“学生与开发者成长计划”，进一步降低 AI 创新的成本，让更多创造者能够依托高质量的信息获取能力，快速验证想法、构建应用，并参与到下一代 AI 生态的构建中。“AI 的发展不仅依赖模型能力的突破，更需要广大开发者将技术转化为真实应用。”薛冠群表示，“我们希望借助这一计划，让全球在校学生和开发者便捷

 如今线上题库、网课、自适应习题工具早已普及，但教育行业质量、成本、规模难以兼顾的“不可能三角”依然尚未打破。同一班级里，有的学生一点就通、有的卡在基础概念、有的陷入思维误区，统一授课很难适配所有人；一对一的个性化教育高度依赖优质师资，成本高、难推广。教育AI工具尽管简单易用，但多停留在内容分发、答题响应与习题推荐的表层逻辑，无法看透学生内在心智变化和潜在需求。教育行业亟待从简单的“答题响应”，走向能模拟、读懂学生内心思考过程的“心智仿真”新阶段。在2026第9届世界人工智能大会（WAIC）期间，由天立国际、启鸣达人、天立启鸣AI研究院、AIII人工智能国际研究院，以及北京航空航天大学国际创新研究院联合打造并发布的《世界模型驱动的教育AGI白皮书——从认知仿真到教育智能体的范式跃迁》（以下简称：《白皮书》），提出以认知世界模型为核心技术底座，通过工程化的认知仿真能力破解教育“不可能三角”，为下一代智能教育基础设施提供理论锚点与落地路径。天立国际首席科学家、启鸣达人公司CTO、天立启鸣AI研究院院长、北京航空航天大学国际创新研究院基础教育通用人工智能实验室首席科学家兼联合主任刘志毅在论坛

OpenAI is deploying a Health feature inside ChatGPT, giving users the option to connect Apple Health data and medical records to the chatbot. Logged-in users aged 18 and older can access it now on web and iOS, across the Free, Go, Plus, and Pro tiers. Users link Apple Health and, where supported, records from US […] The post OpenAI pushes ChatGPT into patient health records appeared first on AI News.

Weeks after Anthropic's latest toe-to-toe with the US government, and days after an OpenAI security incident that dominated tech industry discussions, Anthropic on Thursday released its newest model, Claude Opus 5. The company said in a release that Opus 5 "comes close to the capabilities of Claude Fable 5 in many domains" and is much […]

The newest way to buy enterprise AI agents from OpenAI does not involve buying anything online. OpenAI Presence, announced on July 22, is a managed product delivered through a limited general availability programme, and the company states plainly that it is not yet available as a self-serve product. Deployments are led by OpenAI’s own Forward Deployed Engineers […] The post OpenAI Presence sells enterprise AI agents with engineers attached appeared first on AI News.

On this episode of Uncanny Valley, we dive into accusations that China’s Moonshot AI stole from  Anthropic, and how the US Army needs to cut back on AI use.

Models are improving quickly, but the cheaper options are often good enough.

Meta is upgrading its AI chatbot with new productivity features in a bid to compete with rivals like Gemini, ChatGPT, and Claude. The update will allow Meta AI to tap into your calendar to help you plan events and generate daily briefings, as well as perform in-depth research that you can steer as it progresses. […]

We spoke with several cybersecurity researchers, who look for unknown vulnerabilities and develop tools to exploit them, about how OpenAI’s and Anthropic’s guardrails affect their work.

Chinese AI lab Moonshot’s open model Kimi went viral this week for reasons that had less to do with the model itself and more to do with how the U.S. AI industry reacted to it. Meanwhile, an unreleased OpenAI model wandered outside its test environment and ended up connected to a real security breach at Hugging Face — a reminder […]

Two dozen companies and organisations signed an open letter urging US policymakers to protect open-weight AI models. The letter, published today (PDF), carries signatures from a list that spans direct commercial rivals and organisations with little obvious overlap in business model: Meta, Microsoft, Nvidia, IBM, Dell Technologies, CrowdStrike, Palantir, ServiceNow, Hugging Face, Perplexity, Mistral, Andreessen […] The post Meta, Microsoft, Nvidia, IBM, and others back open-weight AI appeared fir

The acquisition brings Poke’s conversational style and interaction model to Cognition’s coding agent Devin, reflecting a growing belief that how AI assistants interact with users is as important as the models powering them.