Rohan Paul@rohanpaul_ai · 7天前49

AI CXOs are the next enterprise software war. Fairgen just launched "AI Chief Insights Officer", an AI research system that turns real consumer data into queryable simulated respondents for super fast brand, product, pricing, and ad decisions. Fairgen thinks that many early decisions do not need a full $5K to $200K study if teams can get a reliable directional signal in 20 minutes. So Fairgen Twin is meant to behave like one real respondent, built from interviews, surveys, transactions, reports, and panel data rather than from a generic chatbot persona. Fairgen’s answer is daily fresh data collection, plus a 6-dimension quality gate checking logic, fidelity, tone, plausibility, engagement, and numerical coherence.

译Fairgen 推出“AI Chief Insights Officer”，其核心产品 Fairgen Twin 基于每月 10 万次真实访谈为每位消费者构建 1:1 数字孪生。用户可筛选特定人群，在 20 分钟内完成定价、概念、广告测试并生成完整分析报告，替代传统 5 千至 20 万美元的研究。数据来源包括访谈、调查、交易、报告和面板数据，而非通用角色。系统通过 6 维质量门控（逻辑、忠实度、语调、合理性、参与度、数值一致性）确保输出质量。欧莱雅、T-Mobile 等品牌已使用四年，现正式向公众开放。

Thariq@trq212 · 7天前33

I'll be talking more about Claude Tag with @petergyang and at AIE with @_catwu. Let me know if there's anything you'd like us to dive into more!

译我将与 @petergyang 以及在 AIE 上与 @_catwu 进一步讨论 Claude Tag。 如果你们有任何希望我们深入探讨的内容，请告诉我！

ClaudeDevs@ClaudeDevs · 7天前72

Claude Tag is the next evolution of agents. It's a proactive, multiplayer agent with memory and identity, built on top of Claude Code. Learn more about how Claude Tag works and best practices for using it in this deep dive.

译Claude Tag 是智能体的下一代进化。它是一个主动的、多玩家智能体，具有记忆和身份，构建在 Claude Code 之上。了解更多关于 Claude Tag 的工作原理以及使用它的最佳实践，请看这篇深度解析。

Midjourney@midjourney · 7天前67

Two quick updates in image world. Try adding --preview to your prompt for a early peak at V8.2 aesthetics & personalization. We've also updated our big batch draft mode to work with --sref random so you can explore style space 24x faster than before. Enjoy!

译Midjourney 带来两项更新。一是加入 `--preview` 参数可提前体验 V8.2 的美学与个性化效果；二是此前在 V8.1 推出的大批量草稿模式（生成 24 张低分辨率图，价格仅为标准 4 张的一半，点击 "Vary" 可升级为全分辨率）现在支持搭配 `--sref random` 使用，探索风格空间的速度比之前快 24 倍。

Rohan Paul@rohanpaul_ai · 7天前50

i10X just introduced Superagent, an AI Chief of Staff built to turn business goals into finished work. By giving an AI agent access to 100+ tools, cloud execution, research, writing, prospecting, decks, SEO, and human approval gates. Most AI tools still depend on you to break work into pieces, move files between apps, copy outputs, check quality, and decide what happens next. A Superagent tries to absorb that coordination layer. It plans the steps, gathers information, uses connected tools, produces the asset, and keeps running in the cloud even when you are not sitting there babysitting the process. The important design choice is the approval gate. For low-risk work, it can execute quietly; for consequential actions, such as sending an email or publishing a page, it asks first, which turns autonomy into supervised autonomy rather than blind delegation.

译i10X推出Superagent，号称世界首个AI首席幕僚。用户输入业务目标，即可部署智能体团队端到端处理销售、内容创作和SEO。Superagent的核心设计是吸收协调层：自动规划步骤、收集信息、调用100+工具、云端产出成果，无需人工持续监督。关键创新在于审批门控——低风险任务静默执行，涉及发邮件或发布页面等高后果操作时先请求批准，将自主变为受监督的自主，避免盲目委托。

Runway@runwayml · 7天前53

Introducing Agent 2.0 Go from a simple prompt to fully realized marketing briefs and campaign assets right inside of Runway Agent. You can also easily analyze your performance data to improve your creative and scale it across platforms, formats and markets. Agent is being built to become the most capable autonomous agent for real-world work.

译推出 Agent 2.0 从简单的提示词到完整生成的营销简报和活动素材，全都在 Runway Agent 内完成。你还可以轻松分析你的表现数据，以改善你的创意，并跨平台、格式和市场进行扩展。 Agent 正在被打造成为最强大的自主智能体，用于真实世界的工作。

Replit ⠕@Replit · 7天前54

Replit now supports 450+ integrations. No matter what you're building, Replit Agent can wire it up to the tools you already use: payments, messaging, data, CRMs, design tools, analytics, and hundreds more. Just describe what you want. We'll handle the connection.

译Replit 现已支持 450 多项集成。 无论你在构建什么，Replit Agent 都能将它与你已使用的工具连接起来：支付、消息、数据、CRM、设计工具、分析工具，以及数百种其他工具。 只需描述你的需求，我们来处理连接。

ChatGPT@ChatGPTapp · 7天前55

Our plants are chatty. Yours can be too. Give your plants a voice with our build guide: https://github.com/openai/planttalk

译我们的植物很健谈。你的也可以。 通过我们的构建指南让你的植物发声：https://github.com/openai/planttalk

AYi@AYi_AInotes · 7天前62

我深度拆解白毛股神的AI工具就是这个Apodex， 无广，真诚推荐，不好用来骂我！！！ 刚登录准备让它跑个深度研究就睡了， 结果发现弹窗消息免费“内测将于 6 月 29 日 结束， 并正式开始收费，所有在 6 月 28 日及以前注册的 early users，将获得额外 2 周免费试用” 这个真的是个特别好用的工具，强烈推荐宝子们趁着免费期好好用一下～

译AI 工具 Apodex 发布收费通知：免费内测将于 6 月 29 日结束，之后开始正式收费。所有在 6 月 28 日及以前注册的早期用户将额外获得 2 周免费试用。该工具此前被博主深度拆解推荐，适合需要深度研究场景的用户，建议趁免费期尽快体验。

elvis@omarsar0 · 7天前73

Love it that there are more ways to learn with agents. I still like the idea of a dynamic learning hub as we built here: https://x.com/omarsar0/status/2069797099814682638?s=20 More dropping on this soon.

译Google 宣布在 Gemini App 中推出 "study notebooks"（学习笔记本），这是一个交互式学习空间，旨在将好奇心转化为真正的理解。该功能免费且面向全球用户开放，支持所有语言。Elvis Saravia（DAIR.AI）对此表示赞赏，认为这是一种通过 AI 智能体学习的新方式，并提及他之前构建的 "动态学习中心" 概念，后续将有更多相关分享。

xAI@xai · 7天前48

You can now use your SuperGrok and X subscriptions inside T3code

译你现在可以在 T3code 中使用你的 SuperGrok 和 X 订阅。

🚨 AI News | TestingCatalog@testingcatalog · 7天前41

GOOGLE 🔥: Gemini app for macOS will get Computer Use support and will be able to operate other desktop devices remotely. Users will have an option to "Connect another Mac" from the attachment menu on the prompt bar. Closing feature gaps 🤖

译GOOGLE 🔥: macOS 版 Gemini 应用将获得 Computer Use 支持，并能远程操控其他桌面设备。 用户可从提示栏的附件菜单中选择“连接另一台 Mac”。 缩小功能差距 🤖

elvis@omarsar0 · 7天前47

I guess MCP won. Jokes aside, this is super cool from OpenRouter. Just making it easier for devs to run their long-running agents with the right level of intelligence. More of this, please.

译我猜 MCP 赢了。 玩笑归玩笑，OpenRouter 这波操作太酷了。 让开发者能更轻松地用合适的智能层级运行他们的长期 Agent。 请多来点这样的。

Google DeepMind@GoogleDeepMind · 7天前63

Gemini 3.5 Flash now supports native computer use. This built-in tool lets developers build custom agents that can see and take action across browser, mobile, and desktop interfaces. Find out more → https://goo.gle/4f4sNQA

译Gemini 3.5 Flash 现支持原生计算机使用。 这一内置工具让开发者能够构建可跨浏览器、移动和桌面界面进行观察和操作的自定义 AI 智能体。 了解更多 → https://goo.gle/4f4sNQA

Satya Nadella@satyanadella · 7天前42

Today we’re bringing skills to Copilot for Excel, giving teams a new way to scale their expertise across every workbook.

译今天，我们为Excel中的Copilot带来了技能，为团队提供了一种在每张工作簿中扩展其专业知识的新方式。

Google AI Developers@googleaidevs · 7天前73

The Computer Use tool is now available for Gemini 3.5 Flash 🛠️ Build agents that see and take action across browser, mobile, and desktop environments to seamlessly handle long-horizon tasks. Here’s what’s new: - Built-in support for mobile and desktop operating systems - Intent arguments for all function calls - Customizable client-side functions to easily support human-in-the-loop (HITL) takeover - Prompt injection detection and configurable action-level safety policies to fit your app’s needs Learn how you can use 3.5 Flash to automate QA testing, business processes, and more in the blog: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-computer-use-gemini-3-5-flash/

译Google AI 宣布 Gemini 3.5 Flash 的 Computer Use 工具正式可用，支持构建能观察并操作浏览器、移动端和桌面环境的 AI 智能体，可处理长期任务。新特性包括：内置移动与桌面操作系统支持；所有函数调用配备 intent arguments；可定制的客户端函数实现人工接管（HITL）；提示词注入检测及可配置的 action-level 安全策略。可用于自动化 QA 测试、业务流程等场景。

Google Gemini@GeminiApp · 7天前61

Personalize how you study and learn any topic with new study notebooks in Gemini. 📓 Study notebooks can be grounded in your course materials and turn Gemini into an interactive, adaptive learning space with personalized lessons based on your quiz performance to help you learn.

译使用新的学习笔记本，个性化你在 Gemini 中学习和掌握任何主题的方式。📓 学习笔记本可以基于你的课程材料，将 Gemini 变成一个互动、自适应的学习空间，根据你的测验表现提供个性化课程，帮助你学习。

meng shao@shao__meng · 7天前48

家里最累的，往往是那个一直在记事、提醒、安排、照顾所有人的人。 SuperNori 这个方向挺有意思：一个 Proactive Family AI Agent，会先注意到堵车、冰箱快空了、孩子这周在学什么、纪念日有没有撞期，然后再来问你要不要处理。 希望家庭 AI 以后真的能多照顾一下那个总在照顾全家的人。 加入候补名单： https://supernori.heynori.com 你希望家庭 AI 帮你做什么？欢迎转发或留言给 @Nori_FamilyAI #partner

译SuperNori 是一个专为家庭照顾者设计的 Proactive Family AI Agent。与现有AI专注于工作场景不同，它能主动感知堵车、冰箱快空、孩子学习进度、纪念日冲突等家庭事务，再询问用户是否处理。旨在减轻家中“一直记事、提醒、安排、照顾所有人”的成员的负担。目前开放候补名单。

OpenRouter@OpenRouter · 7天前72

Introducing the OpenRouter MCP, live model intelligence right inside your agent Your agent builds and ships, but when it comes to choosing the right model for the right job, it guesses from 6 month old training data Watch it pick, price, and test the right model:

译推出 OpenRouter MCP，将实时模型智能直接嵌入你的智能体 你的智能体负责构建和交付，但在为具体任务选择合适模型时，它只能依据 6 个月前的训练数据来猜测 看它如何挑选、定价并测试合适的模型：

Kling AI@Kling_ai · 7天前11

hҽყ ʏօʊ sҽҽ tԋαƚ UFՕ??

译嘿，你看到那个UFO了吗？？

Chubby♨️@kimmonismus · 7天前25

We spent two years calling things agents that fall over the second nobody's watching. A setup tied to one laptop, one wifi, and one person awake at 1am to restart it when it breaks is closer to a pager than to autonomy. Hyperagent gives every agent its own cloud machine that runs whether you're at the keyboard or asleep, with a real browser and real code execution. Built by the Airtable team. Incredible!

译Hyperagent（由Airtable团队构建）回应了当前智能体依赖单台笔记本、单WiFi及人工守夜重启的脆弱性，为每个Agent分配独立云机器，提供真实浏览器与代码执行环境，确保Agent在离线和无监督状态下持续运行。引用推文对比OpenClaw的频繁崩溃、秘密泄露和持续监控问题，推广Hyperagent：限时注册获$100推理积分，迁移首个Agent再得$500。

Berryxia.AI@berryxia · 7天前11

这几天有空把这个制作流程开源给大家，看看搞一个skills什么都行不行！ 如果你感兴趣就一键三连啊～

译个人开发者Berry Xia在X上宣布，计划将其制作流程开源，并询问能否用该流程生成“Skills”（可能指Claude等AI Agent的技能包）。其此前曾表示“太卷了……先做个垃圾出来”，暗示这是一个快速实验性项目。感兴趣者可“一键三连”关注后续。目前未公布具体模型、版本或参数细节。

向阳乔木@vista8 · 7天前62

腾讯出了专门给Agent用的邮箱，快去抢注你的邮箱名... 注册以后，会有一段提示词，发给Codex或你的其他Agent来完成cli设置。 地址见评论区

Chubby♨️@kimmonismus · 7天前63

Fable 5 is back - and now there’s video proof. Not just showing up in the model selector. People are actually using the model again. We are so back.

译Fable 5 回来了——现在有视频证据。 不只是出现在模型选择器中。人们真的又开始使用这个模型了。 我们回来了。

Alibaba Cloud@alibaba_cloud · 7天前37

Securing AI Agents on Alibaba Cloud: The Constraint Infra ️ Solve Agent chaos with a robust governance layer: ✅ Dynamic Control: Hot-update Prompts/rules via Nacos. ✅ Granular Governance: Token limits & multi-agent security. ✅ Proven in Prod: StarOps SRE Agent runs high-risk tasks safely within these boundaries. ✅ Self-Evolving: Rules iterate via AgentLoop data flywheel. Build safer, smarter Agents! 🚀 https://int.alibabacloud.com/m/1000414834/ #AI #AlibabaCloud #Nacos #Higress #StarOps #AgentLoop

译阿里云发布面向AI智能体的约束基础设施（Constraint Infra），提供治理层解决Agent混乱问题。核心能力包括：通过Nacos热更新提示词与规则实现动态控制；支持token限制及多智能体安全的细粒度治理；已在生产环境验证，StarOps SRE智能体在该边界内安全运行高风险任务；通过AgentLoop数据飞轮驱动规则自我进化。

PixVerse@PixVerse_ · 7天前30

A better way to make virtual livestreams feel interactive. Let the audience choose a look, switch the character in seconds, and post the result as a short video. RT + Follow + Reply = Free Access + 150 credits in DMs.

译让虚拟直播更具互动性的更好方式。 让观众选择外观，几秒内切换角色，并将结果发布为短视频。 转发 + 关注 + 回复 = 免费使用 + 150 积分（私信发放）。

Chubby♨️@kimmonismus · 7天前29

This began as a short film concept that felt unrealistic to produce with traditional tools. My initial thoughts were incredibly vague, but I tried out the new Vibe Create workflow on the Dreamina Octo web platform and the final output came out really good. Here is the final piece:

译这最初是一个用传统工具制作起来不现实的短片构思。 我最初的想法非常模糊，但在Dreamina Octo网页平台上尝试了新的Vibe Create工作流，最终输出效果非常好。以下是最终成片：

Alibaba Cloud@alibaba_cloud · 7天前50

Another exciting milestone for #HappyHorse 1.1! #Fal has joined the growing ecosystem, bringing advanced #AIVideo creation to creators everywhere; unlocking richer storytelling, native audio, and greater creative control.

译Happy Horse 1.1 的又一个激动人心的里程碑！ #Fal 加入了不断壮大的生态系统，为创作者带来先进的 #AIVideo 创作；解锁更丰富的叙事、原生音频和更强的创作控制。

Alibaba Cloud@alibaba_cloud · 7天前49

More ways to create. More ways to inspire. It's fantastic to see #Picsart making #HappyHorse 1.1 available to creators, expanding new opportunities for visual storytelling and video generation.

译更多创作方式。更多灵感来源。 很高兴看到#Picsart将#HappyHorse 1.1提供给创作者，扩展了视觉叙事和视频生成的新机会。

Alibaba Cloud@alibaba_cloud · 6月25日39

Stop shooting. Start creating with #HappyHorse 1.1. Produce captivating brand and commerce videos with intelligent storytelling, expressive motion, refined details, and immersive sound experiences. Built for creators who want speed without compromising quality. Launch Offer: Save 40% https://int.alibabacloud.com/m/1000414698/ #HappyHorse #AlibabaCloud #ModelStudio #GenerativeAI

译停止拍摄。用 #HappyHorse 1.1 开始创作。 打造引人入胜的品牌与电商视频，具备智能叙事、富有表现力的动作、精细细节和沉浸式音效体验。 专为追求速度又不愿牺牲质量的创作者打造。 首发优惠：节省 40% https://int.alibabacloud.com/m/1000414698/ #HappyHorse #阿里云 #ModelStudio #生成式AI

向阳乔木@vista8 · 6月25日49

谷歌六月参考AK的LLM wiki思路推出了一个规范：Open Knowledge Format 核心思路：用一组 Markdown ，加 YAML frontmatter，把知识变成可读、可版本控制、可被 Agent 直接消费的文件包。 让Codex读取这篇文章，它把codex做过的项目整理成了OKF格式，添加到系统记忆索引中。 文章见评论区

译谷歌于六月推出Open Knowledge Format（OKF）规范，参考AK的LLM wiki思路。核心是用Markdown加YAML frontmatter将知识组织为可读、可版本控制、可直接被AI智能体（Agent）消费的文件包。示例中，Codex读取该文章后，将自身做过的项目整理为OKF格式并添加到系统记忆索引中。文章见评论区。

Alibaba Cloud@alibaba_cloud · 6月25日36

One dress. Three occasions. Zero compromise on consistency. With HappyHorse 1.1, transform a single product into multiple lifestyle narratives while preserving subject identity, styling details, and visual coherence across every scene. Perfect for fashion brands looking to scale creative production without reshoots. Limited Launch Offer: 40% OFF via API https://int.alibabacloud.com/m/1000414698/ #HappyHorse #AlibabaCloud #ModelStudio #GenerativeAI

译一件连衣裙。三个场合。零妥协的一致性。 借助 HappyHorse 1.1，将单一产品转化为多个生活方式叙事，同时保持主体身份、造型细节和跨场景的视觉连贯性。 适合希望在不重拍的情况下扩大创意生产的时尚品牌。 限时发布优惠：通过 API 享 40% 折扣 https://int.alibabacloud.com/m/1000414698/ #HappyHorse #阿里云 #ModelStudio #生成式AI

AYi@AYi_AInotes · 6月25日71

卧槽这个必须分享，一个开源工具，让你用免费 API 密钥池跑出企业级路由的效果， 等于是零成本撸10亿+免费LLM Token，要把把付费网关干碎的节奏， 对比一下：高容量令牌和企业路由，0， 原理很简单，它是个路由框架，不是卖 API 的， 你需要自己去各厂商申请免费密钥，然后填进配置， 工具自动帮你做负载均衡和自动故障切换， 30 秒能跑起来：克隆仓库，配好密钥，把应用指向本地端点，完事， 免费额度用满、用稳，不用自己写回退逻辑， 项目几周前刚发布，现在入坑还能直接给作者提改进意见， GitHub 链接放评论区 👇 有用的记得给仓库加星。

译一款开源路由框架（非API售卖），让用户自行申请各厂商免费API密钥，通过配置实现自动负载均衡与故障切换，从而零成本使用10亿+免费LLM Token。操作极简：克隆仓库、填入密钥、将应用指向本地端点，30秒即可运行，无需手写回退逻辑。项目几周前刚发布，作者开放改进建议，GitHub链接见评论。

Alibaba Cloud@alibaba_cloud · 6月25日25

Sunflowers in the Wind The adult world is exhausting. Old Zhou, a taxi driver of twenty years, had grown numb to it all — the bridges, the sea breeze, the city lights. He was ready to quit. Then a little girl got in. She didn't rush him. She simply gazed at everything — the bridge, the park wind, the river clouds, the purple evening sky — savoring it all like a quiet goodbye. It was her last look at the world she loved, before surgery. Innocent yet brave. Gentle yet determined. In that moment, Old Zhou realized: the ordinary life we've grown tired of is the tenderness others desperately wish to keep. One sunflower. Two hearts healed. The next morning, light fell over a field of flowers. Old Zhou drove on — rekindled. Ordinary life, gently lit again by warmth and hope. Happyhorse CPP Qingshan Yingling studio × Happyhorse short film “Sunflower” 🌻 HappyHorse 1.1 is now available! HorsePower AI Cinema Awards Global Competition now open! More info can click: https://int.alibabacloud.com/m/1000412436/

译阿里云发布 HappyHorse 1.1 版本，同时启动 HorsePower AI Cinema Awards 全球竞赛。短片《向日葵》作为宣传内容，讲述了出租车司机老周与重病小女孩之间的温情故事，展现该平台在情感叙事与 AI 创意生成方面的能力。

Berryxia.AI@berryxia · 6月25日74

PaddleOCR的PP-OCRv6终于上Hugging Face了。 这次不只是精度又提升，还一次性加了transformers和ONNX Runtime两个后端。 意味着你现在可以用更统一的API，在不同推理框架之间无缝切换，而不用改太多代码。 PaddleOCR一直以来都是工业界用得最多的开源OCR方案之一，这次上HF + 多后端支持，等于把门槛又拉低了一大截。 尤其是想在transformers生态里直接用高性能OCR的人，这次可以直接上手了。 从之前的Unlimited-OCR到这次PP-OCRv6，国内团队在长文档和实用OCR方向上确实在持续迭代，而且越来越注重工程可用性。 地址见评论区👇

译PaddleOCR 的 PP-OCRv6（对应 PaddleOCR 3.7）正式上线 HuggingFace，精度进一步提升，并新增 transformers 和 ONNX Runtime 两个推理后端。用户可通过统一 API 在不同后端之间无缝切换，无需大幅修改代码。PP-OCRv6 是工业界广泛使用的开源 OCR 方案，此次上架 HF 并支持多后端，降低了工程接入门槛，尤其利好希望在 transformers 生态中直接使用高性能 OCR 的开发者。

Luma@LumaLabsAI · 6月25日31

A guide. A descent. One of the oldest stories we have, rebuilt frame by frame. Dante's Inferno by @NinaSabinaAI . Made with Luma.

译一段指引。一次坠落。 我们最古老的故事之一，逐帧重建。 《但丁的地狱》由 @NinaSabinaAI 创作。使用 Luma 制作。

Artificial Analysis@ArtificialAnlys · 6月25日61

Agentic knowledge work can take frontier models over 20 minutes per task, as measured in AA-Briefcase, our new benchmark Last week we released AA-Briefcase, our proprietary agentic knowledge work benchmark testing models on long horizon tasks built by industry experts. AA-Briefcase requires models to build deliverables such as financial models, board presentations, and design mock-ups in the context of realistic multi week projects. One of the key metrics we measure in AA-Briefcase is average time per task. This is calculated using evaluation token usage, representative model output speeds, and tool execution time recorded during evaluation. Key time per task takeaways from AA-Briefcase: ➤ Claude Opus 4.8 is the highest-scoring available model, but it is also one of the slowest, taking ~23 minutes per task on average ➤ Several GPT-5.5 reasoning variants lie along the Pareto frontier of AA-Briefcase Elo vs. Time per Task, including medium, high, and xhigh. GPT-5.5 (xhigh) in particular stands out as one of the most efficient top-performing models, using around half the time per task of Opus 4.8 (11 minutes) while ranking top 5 on the overall AA-Briefcase Elo ➤ GLM-5.2 also sits on the Pareto frontier, scoring 1261, ahead of GPT-5.5 (xhigh, 1159) but also taking more time per task (16.3 minutes). It is also the top-performing open weights model on AA-Briefcase, with MiniMax-M3 the next best at 1113 ➤ If Claude Fable 5 were still available, it would likely take around 28.5 minutes per task: while it was live, we measured ~91 output tokens per second, ~3.1 minutes of tool execution time per task, and ~139,000 output tokens per task ➤ Time spent on tool calls and execution accounts for only ~12% of the total time, with the remaining amount explained by output verbosity, turn usage, and inference speed

译Artificial Analysis 发布 AA-Briefcase 基准测试，测试模型在多周项目语境下生成财务模型、董事会演示等交付物。关键结果：Claude Opus 4.8 平均每任务 23 分钟，得分最高但最慢；GPT-5.5 (xhigh) 仅 11 分钟，效率最高且 Elo 前五；GLM-5.2 得 1261 分耗时 16.3 分钟，为开源模型最佳；MiniMax-M3 得 1113 分。已下架的 Claude Fable 5 约需 28.5 分钟。工具调用仅占耗时 12%，其余由输出冗余、回合数和推理速度决定。

Luma@LumaLabsAI · 6月25日31

The toilet paper is gone. The bathroom is empty. The kid is too quiet. Zozo, a claymation short by Eran May-Raz. Made with Luma .

译卫生纸没了。浴室空了。孩子太安静了。Zozo，一部由Eran May-Raz创作的粘土动画短片。使用Luma制作。

🚨 AI News | TestingCatalog@testingcatalog · 6月25日60

Google Flow Agent can now use Google Maps Street View grounding to generate images and videos. Only works with US locations at this moment. > "Your Google Flow Agent can now generate images and videos grounded in Google Maps Street View, giving your scenes real-world details and accuracy, whether you are placing characters in a local scene, stylizing a city landmark, or reimagining your favorite neighborhood." > "To use the feature, enable Agent mode and use a named location or a street address in your prompt. This feature is available globally, but currently only for US locations."

译Google Flow Agent 新增 Google Maps Street View 实景锚定功能，支持根据用户提示中的地名或街道地址，为图片和视频生成提供真实世界细节。用户需启用 Agent 模式并在提示词中包含具体地点。目前该功能仅限美国街景位置可用。

elvis@omarsar0 · 6月25日43

This is a cool release from Google. I've started to rely a lot on computer use for agentic loops/long-running tasks, and I'm in need of more capable computer use models while keeping costs low. Not sure about efficiency of Gemini 3.5 Flash, but I appreciate alternatives.

译这是 Google 的一次酷发布。 我开始在智能体循环/长时间运行任务中大量依赖计算机使用，并且需要更强大的计算机使用模型，同时保持低成本。不确定 Gemini 3.5 Flash 的效率如何，但我感谢有替代方案。