Claude 表格白板会议全套上线：直接叫板微软与谷歌办公套件
Anthropic has launched Claude Sheets and Claude Whiteboards in beta and made Claude Data, previously called Claude Dashboards, generally available on all paid plans. It also introduced Claude Meetings, which puts Claude into video calls so it can start on tasks while people are still talking.
The aim, Anthropic says, is to keep documents, boards, sheets and dashboards inside Claude, where it can follow changes live, instead of copying output into other tools.
Claude Sheets builds spreadsheets with tabs, formulas, formatting and drop-downs, often from a single prompt such as a campaign tracker or hiring budget. Claude edits directly in cells and outlines each change so teams can track what it did. Anthropic says the feature set is simple for now; files can be exported to.xlsx or CSV, or moved to partner tools including Airtable, Asana, monday.com and Smartsheet.
Claude Whiteboards turns a spoken or messy plan into a shared board with a timeline, owners and dates. In a retro, Claude can group notes by theme and draft action items, and the finished board can become a doc or deck, or move into Lucid, Miro or Vani by Zoho.
Claude Data connects to sources such as Snowflake and Salesforce and can use definitions certified by a data team to build dashboards, data stories or narrated short videos. Anthropic also upgraded interactive connectors so partner apps can be used inside Claude.
With Claude Meetings, a team adds Claude to a call as a visible participant. It can draft the plan being discussed, map a timeline on a whiteboard, answer questions in the meeting chat and pull figures from connected work apps. Notes, action items and other material it produces are then available in claude.ai.
Claude announces itself in the meeting chat, will not join calls with external guests unless an admin authorises it, and leaves automatically if an external guest joins. Meetings is in beta on Team and Enterprise plans and supports Google Meet at launch.
Sheets, Whiteboards and Data work on web, desktop and mobile; Data needs a connector such as a warehouse or CRM. On Team plans all three are on by default. On Enterprise plans Sheets and Whiteboards start off, and Claude Data switches on by default on 19 November 2026.
Anthropic is moving Claude from a chat window that produces files into a workspace where the files live. That puts it in more direct competition with Microsoft 365 and Google Workspace for everyday office work, while still offering exports and partner integrations for teams that are not ready to switch.
ChatGPT 变身商家平台：12 亿周活用户面前开店，还能挂广告
We’re introducing Businesses in ChatGPT in beta, giving businesses a new way to connect with the 1.2 billion people who use ChatGPT each week. Through Business Hub, businesses can claim their business, manage their information, and see insights on how people engage with them on ChatGPT. In addition, we are adding a powerful new way for businesses to interact with users via Business Agents, starting with eligible ecommerce and online service businesses in the U.S.
In Business Hub, companies can claim their organization in ChatGPT and, once verified, update key details and add self-service product feeds with pricing and availability. This gives businesses a new way to provide information to ChatGPT, helping people explore products and services using the most accurate and up-to-date details.
Any eligible business located where ChatGPT is available can claim their business in ChatGPT through Business Hub. Existing OpenAI Enterprise customers and advertisers can work directly with their OpenAI representative to claim their business. Shopify merchants can connect their store from Business Hub or the Shopify admin to streamline verification. Eligible products from Shopify merchants are already integrated into ChatGPT through Shopify Catalog.
Business Hub provides aggregated insights into how people engage with a business in ChatGPT. This includes the number of mentions and website clicks a business receives. Businesses can also explore discovery topics—common themes across conversations where their website is cited. These insights can help businesses understand how they’re represented and find opportunities to improve the information they share.
We’ve designed these insights to protect user privacy. We don’t share users’ conversations or Business Agent chats with businesses. Insights are aggregated and are only available when there are enough conversations or interactions to meet certain thresholds.
After claiming their business in ChatGPT, eligible organizations can build, test, and launch a Business Agent—a new way to answer questions and help people explore their offerings.
When a business appears in a model response, people can click to see the business’s ChatGPT profile. If a Business Agent is available, customers can choose to start a conversation with the business to explore its offerings and ask follow-up questions. For example, someone comparing products could ask about the differences between two options and which features fit their needs. A Business Agent could help them explore those questions using the business’s information.
Businesses can also sponsor their Business Agent and use it as the destination for ChatGPT Ads to reach more potential customers. Sponsored placements are clearly labeled and separate from ChatGPT responses.
As people turn to AI to discover products, compare options and make decisions, ChatGPT has become the platform for people and businesses to connect through AI. Today’s announcement is an early step in bringing businesses more deeply into ChatGPT.
关键词mentions and clicks
谷歌 Gemini 4 Carbon 全面开放：最强模型主攻长程智能体与网络安全
Today, we’re making Gemini 4 Carbon available to developers, enterprises, scientists, and consumers. Carbon is our strongest model with frontier performance in cybersecurity capabilities and long-horizon agentic workflows across complex knowledge work, software engineering, and scientific reasoning.
Two weeks ago, we released Carbon to a set of cybersecurity defenders via our Fairwind Program. Following our recent launch of Gemini 4 Argon, Carbon pushes the frontier even further. Argon remains the ideal choice for the majority of software engineering and knowledge work, while Carbon is built for tasks that test the limits of a model’s intelligence like complex, multi-step coding and debugging and open-ended scientific research and discovery — using more tokens than Argon to explore thoroughly, reason deeply, and run extended verification loops.
Across independent benchmarks designed to test model performance in real workflows, Carbon pairs frontier accuracy with practical production economics. It ranks #1 on CursorBench 4.0, Perplexity’s WANDR, and Zapier’s AutomationBench at more than 2X lower cost per task compared to the next best external model.
To put Carbon’s capabilities to work on larger and complex projects, we’re previewing `/swarm` in Google Antigravity, turning a single prompt into a coordinated team of agents you can track in real time. Developers can assign Carbon to roles that demand the deepest reasoning, like architectural design, coordination, and auditing, while pairing it with faster, lighter models across parallel isolated workspaces and adversarial verifiers that test every change.
Gemini 4 Carbon is available via the API in Google AI Studio and Gemini for business for all paying developers and enterprises, at the same introductory price as Gemini 4 Argon ($2 per million input tokens and $10 per million output tokens). It is also rolling out to all Google AI Ultra subscribers in Google Antigravity, the Gemini app, and AI Mode in Search.
Gemini 4 Carbon for cyber delivers our strongest cybersecurity capabilities to date, giving security teams the end-to-end intelligence they need in protecting critical software. When accessed through the Fairwind Program, Carbon leads on key benchmarks across finding (SEC-bench Pro), exploiting (ExploitGym), and patching (CWE-bench v1) vulnerabilities: inspecting software, tracing how flaws can be chained together across systems, and generating verified root-cause patches.
Code verification: Researchers built a new agentic framework for formal verification called AlphaProof Code and put it to work on a major security challenge: mathematically proving the actual code underlying the cryptographic protocol of Signal. Using Carbon alongside other models, our researchers translated security goals to mathematical theorems. AlphaProof Code agents then used Carbon to generate millions of lines of code in Lean, a computer language for writing mathematical proofs, helping them prove key security properties in weeks — a task that historically takes expert teams years.
阿里 Qwen4 即将发布：Max 预览版先上 Token Plan，1 折体验
11月11日，阿里巴巴宣布，全新一代千问大模型Qwen4即将发布，Qwen4-Max预览版已首发上线阿里Token Plan。千问办公、Qoder、千问App和QwenStudio等产品也已接入新模型。
据介绍，Qwen4将进一步提升知识理解、世界感知、编程及智能体能力。在复杂编程、专业办公、前端设计和专业软件操作等实际任务测试中，新模型已完成相关测试。不过，阿里巴巴尚未在本次信息中披露Qwen4正式发布时间、具体参数及完整型号阵容。
目前，用户可通过阿里Token Plan体验Qwen4-Max预览版，活动期间最低折扣为1折。Token Plan是阿里云推出的大模型订阅服务，可使用Credits抵扣不同模型调用及智能体工具相关费用，并支持文本和多模态模型。
此前，阿里云于11月9日调整Token Plan个人版订阅方案。其中，Lite、Essential和Standard三档套餐月费保持不变，月度额度提升约74%至78%；Pro套餐额度由18万Credits提高至20万Credits，月费由499元降至312元。此次调整同时覆盖存量订阅用户，新额度自11月9日起陆续生效；Pro存量用户当前订阅周期价格不变，新价格从下一订阅周期开始执行。
双11 Token 成爆款：大模型厂商集体上天猫打五折
今年双11，天猫的品牌名单里，多了一批特殊的商家。
智谱、MiniMax、千问办公、阿里云、Qoder。
这些平时忙着训练模型、比拼技术的AI大厂，到了11月11日，竟然集体打5折。
MiniMax三款月卡全部5折，最低25元起；阿里云两款Token Plan月卡5折起；智谱周卡5折起；千问办公基础版会员首月低至3折，Qoder月卡更是降到了19.9元。
打开淘宝App搜索“Token”，进入天猫AI空间站，就能挑选不同厂商的AI服务套餐，享受双11当天的限时优惠。
它们卖的东西，叫Token。
这是一种看不见、摸不着，买完无需等快递的数字算力。过去只存在于程序员和开发者账单里的技术服务，如今也贴上了双11的促销标签。
11月2日晚，千问办公CEO陈宇森带着一款AI办公会员月卡走进李佳琦直播间，原价28元，当晚只卖1分钱。
开卖10分钟，售出5万份；半小时，累计售出近10万份，对应约2万亿Token的AI服务额度。
9月2日，智谱率先入驻天猫开设官方旗舰店；次日，天猫正式上线“AI空间站”Token充值中心；10月13日，MiniMax也紧随其后，开出品牌首家天猫官方旗舰店。
到了双11，阿里云百炼、千问办公、Qoder、智谱、MiniMax、月之暗面、阶跃星辰等品牌的模型与服务，已经齐聚天猫AI Token专区。
据天猫披露，10月20日双11开售首日，Token品类成交额较上月增长450%。此前，智谱入驻天猫后48小时内，品牌搜索量也曾环比增长50倍，并带动相关AI订阅产品成交。
从开发者圈子里的技术指标，到普通消费者购物车里的商品，Token在今年双11完成了一次身份蜕变。
Grok 4.8 预训练完成：xAI 扔掉 JAX/SGLang，全栈手写 C 重写
Elon Musk gave a status update on Grok 4.8 on 10 November 2026, and the interesting part is not the model — it is what xAI tore out to build it.
In Musk’s words: “We have dumped JAX/SGLang and rewritten the entire RL and inference stack from scratch in C/C++, which took some extra time.” JAX and SGLang are standard infrastructure for training and serving large models. Replacing both with hand-written C is not a tweak.
He described the result as “C with a little C++, but not too much” with some assembly, and added that the team writes its own headers and calculates its own time including leap seconds. His summary: “This is tight code.”
“Training (aka pre-training) is done.” He added that a lot of cleanup was needed on the SFT data to get it mostly right.
“Now, 4.8 is in RL and the evals are starting to climb.”
A limited Bot-user rollout later that week, to see how it goes. No wider date.
The model is 2.5 trillion parameters trained with the new C++ stack, per the September roadmap.
Asked how Grok 4.8 compares with Claude Opus 5.5, Musk said it was not as good and called Opus 5.5 a “masterpiece”. He said Grok 5 should exceed it, and that this might happen in two or three months.
A chief executive naming a competitor’s model as better than his own unreleased one is rare enough to be worth noting. It also sets a public marker: Grok 5, past Opus 5.5, by roughly February.