搜索“Agent”

Mercury 2 for Search: Fast enough to run a hundred times per query Mercury Voice is a diffusion LLM (dLLM) tuned to power voice agents. It reasons, calls tools, and follows long system prompts while keeping latency low enough for natural conversation. Mercury 2.5 is the most capable diffusion LLM on the market. It runs at 1,107 tokens/sec and offers a 40% increase in intelligence over Mercury 2, comparable to cost-optimized frontier models. Sep 8, 2026 Product Mercury 2 for Search: Fast enough t

匹配字段正文

xAI’s Grok 4.7 is now available on Amazon Bedrock, adding a frontier model built for coding, long-running agents, and knowledge work to the Bedrock model catalog. It offers a 500K token context window and supports configurable reasoning effort at four levels: low, medium, high, and xhigh. Grok 4.7 is served on the bedrock-runtime endpoint through cross-Region inference profiles, and it supports the Responses, Chat Completions, and Converse APIs. According to xAI, it’s their most capable model fo

匹配字段正文

rch.text-copy'); sessionStorage.removeItem('facets.search.filters'); Sep 28, 2026 NVIDIA OpenShell 0.1.0 provides an open-source runtime that enforces which systems and data an AI agent can access without rewriting the agent. The runtime combines sandboxed execution, controlled service access, credential management, and formal policy analysis to restrict API operations and protect credentials outside the agent workload. Organizations including Cadence, Slack, and Gecko Robotics are adopting Open

匹配字段正文

nd run it in a separate Python environment with the prerequisites listed below. Harbor: run a reference solution Harbor can load task directories from a Hub repository. The oracle agent runs the task's reference solution, then the verifier scores the result. It does not call a model. The viewer shows the task's reward, verifier output, and logs. This checks the task and its reference solution before you try a model agent. Verifiers: run a model on the same task The Harbor integration of verifier

匹配字段正文

, effort may not be applied consistently. Also, the environment can affect the results of the evaluation: leftover state from an earlier trial (a file, a git history) can hand the agent the answer Adversarial sampling Model capability is jagged. If you pick cases because today's model fails them, you are sampling the valleys of one model's capability surface (Figure 2). The evaluation can end up measuring that model's failure fingerprint rather than what is intrinsically hard or valuable for you

匹配字段正文

try it out yourself in Google’s AI tools. Gemini 3.8 Flash is our most intelligent workhorse model, delivering significant improvements from 3.7 Flash across software engineering, agentic tasks, and critical, multistep reasoning in specialized domains. These performance gains stem from a core design choice: 3.8 Flash works harder. On complex tasks, it exhibits greater diligence — executing extra reasoning steps and calling tools iteratively to land a more accurate output. Builders have been crea

匹配字段正文

s or to an Opus-only control group, across the same workloads and users. Cost per session fell 57% (±19 ppts), from $15.36 to $6.63. To score the accuracy, we verified whether the agent finished the work, whether the right answer was reached, and whether the user had to make any corrections on the next turn. FireRouter with Opus scored 78.7% of graded turns against 80.2% for Opus-only, or 98.1% of its overall accuracy. Most of our traffic came from internal coding work, where open models could h

匹配字段正文

s or to an Opus-only control group, across the same workloads and users. Cost per session fell 57% (±19 ppts), from $15.36 to $6.63. To score the accuracy, we verified whether the agent finished the work, whether the right answer was reached, and whether the user had to make any corrections on the next turn. FireRouter with Opus scored 78.7% of graded turns against 80.2% for Opus-only, or 98.1% of its overall accuracy. Most of our traffic came from internal coding work, where open models could h

匹配字段正文

ode or summarize. It takes unstructured state and returns typed decisions with calibrated probabilities. That makes it a natural fit for the thousands of small judgments inside an agent loop: which model to call, whether a command is safe, which passage is relevant, whether the agent is actually done. Every call sends a state (text or JSON) plus a dictionary of typed questions. TypeSafe’s docs define 3 primitives: All questions are evaluated in parallel against the same state in one request. Typ

匹配字段正文

While some retailers, like Amazon (and Adidas , apparently!), are blocking AI agents from making purchases on users’ behalf on their respective platforms, e-commerce platform Shopify has moved in the other direction. On Monday, the company announced that browser-based AI agents can now complete purchases on Shopify merchants’ sites, extending their capabilities beyond just searching for products and adding items to carts. Shopify previously supported WebMCP for its storefronts and carts, allowin

匹配字段正文

托量子物理有望拓展现有人工智能的能力边界，为通用人工智能的发展探索新的技术路线。 △ 图源清华大学官网 在姚宏看来，Q4AI目前正处于产业发展的早期阶段。研究院相关量子研究团队与费米宇宙合作的产业级技术成果，是人工智能领域里具备重要意义的产学研有机结合的实践，期望对下一代人工智能的发展产生深刻影响。 量子怎么进入大模型？ 一类更依赖量子硬件，比如让大模型充当Agent，再去调用量子计算设备完成特定优化任务。 费米宇宙走的是第二条路。 眼下，所谓的「Quantum for AI」，很多工作仍然发生在「量子启发」层面。 费米宇宙方面告诉量子位，海外已出现了一批玩家。 比如SandboxAQ，主要布局后量子密码和量子启发优化；微软Azure Quantum Elements更偏向AI4S和科学仿真；Xanadu则提供量子机器学习科研框架。 国内的企业则两条路并行，技术路线远未收敛。 作为一家基础大模型企业，传统大模型厂商所面向的各类客户均属于费米宇宙的目标客群。与此同时，传统大模型企业也是其服务对象。 接下来6到12个月，费米宇宙计划把现在针对单一基座模型的量子增强能力，进一步抽象成一套通用

匹配字段正文

ery model, every form factor, and every task. Today, we’re announcing our biggest update to Copilot to date, bringing four things together: · Autopilot: proactive and long-running agent built for the enterprise · Code:… pic.twitter.com/W2ClHHkCK3— Satya Nadella (@satyanadella) September 25, 2026 HomeとCodeは、一般提供前のAI機能を試せるFrontierプログラムで今後数週間に順次展開する。Autopilotは9月末にプライベートプレビューの対象を拡大する予定。機能の利用に応じた従量課金と、企業が利用額を管理する仕組みも案内した。 HomeにChatとCoworkを集約⁠、Copilot内でOffice文書の作成⁠・編集も Homeは、Copilotで作業を始めたり、途中の仕事を再開した

匹配字段正文

Today Microsoft announced Autopilot, an always on agent built on OpenClaw The best part of this collaboration is how much @OmarShahine and others at Microsoft have contributed BACK to OpenClaw Read all about the contributions Omar and his team made here: https://t.co/JOXd6sswZ2

匹配字段正文

rch引擎 ：自动接棒，把剩下的实验拆解、跑通、反馈全流程自动化。 最终结果也揭晓了，他们靠这套方法在RoboDojo拿下了第一。 更关键的是速度。 Simate从成立到登顶RoboDojo，前后只用了三个月。 真正的机器人研发，底层穿透了模型代码、数据管线、训练集群、评测环境、真机反馈，以及过去数月甚至数年的实验记录。 归根结底，这依然是「上下文」的问题。Agent需要一个能够调度一切的超级接口。 为此，Simate搭建了一套三层架构：SiPAI ＋ AutoResearch ＋ AI-native Infra。 这是一套原生面向「自动化研究」的可插拔模型框架。 只有这样，Agent拿到研究课题后，才能精确知晓：该改哪个模块？改动会波及什么？最后又该如何验证？ 目前，Simate已经完成了这套基础设施的搭建，成功接入了世界模型、世界动作模型、VLA、VLM等多种主流架构。 无论是人类研究员还是Agent，都能自由组合组件、调整局部实现，并沿用统一的训练与评测流水线。 如果说SiPAI提供了一份静态说明书，那么AutoResearch解决的就是动态信息差： 研究推进到哪了？以前踩过什么

匹配字段正文

echnology aimed at making purchases possible through its AI services. Earlier this year, it introduced the Universal Commerce Protocol (UCP) as an open standard designed to let AI agents interact with retailers across the shopping journey, including checkout. Google said at the time that the technology would allow shoppers to buy eligible products through Gemini and AI Mode using a Google-hosted checkout experience. The company has since expanded UCP with other capabilities, including allowing s

匹配字段正文

ery model, every form factor, and every task. Today, we’re announcing our biggest update to Copilot to date, bringing four things together: · Autopilot: proactive and long-running agent built for the enterprise · Code: build apps with Copilot, hosted inside your company’s tenant · Home: Chat + Cowork together · Office: now fully embedded in Copilot (and Copilot embedded in Office, of course!) Plus, you can invoke Copilot in Teams, and we’re introducing Today, a proactive experience that surfaces

匹配字段正文

hain device called Charm on the way and Meta’s AI glasses next in line, Muse is the new star of the company’s hardware lineup, and Mark Zuckerberg suddenly has one of the first AI agents worth wearing. In today’s AI rundown: Meta’s Connect turns into a Muse takeover Rowan’s Corner: The Rundown team gets together IRL Google’s AI chips are headed to space LATEST DEVELOPMENTS Image source: Meta The Rundown: Meta just introduced a flurry of new AI upgrades for its viral Muse agent at its Connect 202

匹配字段正文

Copilot “super app” last month , Microsoft is officially unveiling it today. The redesigned Copilot app bundles three AI capabilities into a single interface of chat, coding, and agents. As part of the launch, Microsoft is also rebranding Scout, the AI personal assistant it unveiled at Build earlier this year, as Autopilot. This new Copilot app will have three tabs for Home, Code, and Autopilot, with Home combining Copilot Chat and Cowork together. It’s the default landing experience for Copilo

匹配字段正文

库、工具链和通信库等领域的开放范围。 发动机性能再强，最终能发挥多少，还取决于软件能否让模型顺利迁移、运行和优化。 现在，这套配套软件正进一步向开发者开放。 据悉，真武AI芯片已服务20多个行业的650多家客户，蚂蚁、小红书、小鹏汽车等头部企业已经开始使用 T-Head SAIL 。 其中，小红书还基于T-Head SAIL开源代码，开发了模型迁移与算子优化Agent，用来加速生成式推荐模型在真武上的部署。 可以说，平头哥提供基础软件能力，客户则把自己的业务经验写进工具，继续做迁移和优化，芯片背后的生态软实力，也在这样的参与中逐渐积累。 而以阿里为代表的大厂造芯，也正在从交付芯片，走向更广泛的开放共建阶段。 简单说来， T-Head SAIL 可以理解为阿里平头哥围绕真武AI芯片打造的一套类“CUDA”软件栈。 它连接PyTorch等上层AI框架和底层真武硬件，负责模型迁移、编译执行、计算加速和开发调试。 今年7月WAIC上，平头哥宣布启动 T-Head SAIL开源 ，并向开发者提供SDK、驱动、性能分析与调试工具，以及相关技术文档的下载。 两个多月后的云栖大会，团队公布了进一步的开

匹配字段正文

Takeaways Yesterday at Connect 2026, we announced we’re bringing Muse to our AI glasses so you can connect with your agent hands-free throughout your day — plus new features and connectors that make your Muse even better; We introduced Meta VR Glasses: all the magic of VR immersion in a breakthrough pair of glasses you can comfortably wear for hours; We also unveiled our new lineup of AI glasses, including audio glasses, and by the end of this year, we’ll have more than 100 styles of AI glasses

匹配字段正文