Azure

Claude Opus 4.6 上线 Azure Foundry:1M 上下文与 Agent 控制

3分钟阅读

摘要

Anthropic 的 Claude Opus 4.6 已正式上线 Microsoft Foundry on Azure,带来 1M token 上下文窗口(测试版)、最高 128K 输出,以及 Adaptive thinking、Context Compaction 和更细粒度的推理控制等新能力,面向编码、企业知识工作、金融分析和安全等复杂场景。此举的重要性在于,企业可在 Azure 的安全治理与可审计控制平面内部署更强的代理式 AI 系统,同时连接 Foundry IQ、M365 和 Fabric 等数据源,提升业务自动化与高价值知识工作的效率与可信度。

需要Azure方面的帮助?咨询专家

Introduction: why this matters for enterprise IT

随着组织从“copilot-style”的辅助,走向可在业务应用间执行多步骤工作的代理式系统(agentic systems),两项需求成为主导:能力(推理、长上下文理解、工具使用)与 信任(治理、访问控制、可审计性)。如今 Claude Opus 4.6 已在 Microsoft Foundry on Azure 上可用,团队可以在一个面向企业、用于安全且可扩展 AI 工作负载的平台内,部署 Anthropic 最先进的 Opus 模型。

What’s new in Claude Opus 4.6 on Microsoft Foundry

Frontier model capabilities aimed at real workloads

Claude Opus 4.6 面向复杂且高风险的任务场景,定位包括:

  • Large-scale coding(重构、缺陷检测、复杂实现)
  • Enterprise knowledge work(搜索、分析、生成高质量成品文档)
  • Financial analysis(跨申报文件、市场数据与内部来源的分析)
  • Cybersecurity reasoning(检测细微模式与攻击路径)
  • Computer use / UI automation(在应用中导航、填写表单、跨工具迁移数据)

Expanded context and output limits

  • 1M token context window (beta),超过 200K tokens 的部分采用 premium 定价
  • Up to 128K output tokens,支持在单次响应中生成长篇交付物(报告、playbook、多文件代码生成)

Foundry data activation and governance alignment

在 Foundry 内,Opus 4.6 可通过利用 Foundry IQ 来“activate knowledge from everywhere”,访问如 M365 Work IQFabric IQ 以及 web 内容等来源,同时仍保持在基于 Azure 的企业控制平面之内。

New API capabilities co-launching with Opus 4.6

  • Adaptive thinking:动态调整推理深度,在速度与复杂度之间进行权衡
  • Context Compaction (beta):对较早的对话上下文进行总结,以支持长时间运行的 agent 工作流
  • Max effort control:新增更细粒度的控制(在 high/medium/low 之外新增一个等级),用于管理 token 在 thinking、tools 与 output 之间的分配方式

Impact on IT administrators and end users

  • 为工程团队带来更快的交付节奏:可将长时间运行的编码任务委派给 agent,使工程师更多聚焦于评审与架构决策。
  • 更高质量的知识工作:业务用户可更一致地生成符合领域要求的文档与分析——在受监管职能(财务、法务)中尤为有用。
  • 更强的自动化潜力(同时需要更多监督):增强的“computer use”能力提高了端到端工作流自动化的上限,但也提升了对防护栏、日志记录与审批步骤的需求。
  • 成本与容量规划变得关键:1M 上下文与 128K 输出非常强大,但可能推高 token 使用量,并需要围绕何时启用高 effort 推理建立治理机制。

Action items / next steps

  1. 在 Foundry 中 试点 Opus 4.6,建议双轨推进:(a) 编码工作流(repo 重构/评审),(b) 业务工作流(政策草拟、财务摘要)。
  2. 定义治理控制:数据访问范围、工具权限、审计日志,以及对 agent 操作的人在回路(human-in-the-loop)检查点。
  3. 使用 adaptive thinking 与 effort controls 建立 token/成本护栏;仅在确需深度历史的场景中保留 1M-context 运行。
  4. 在扩展到生产工作流之前,于非生产环境中 验证 computer-use agents 的自动化安全性

需要Azure方面的帮助?

我们的专家可以帮助您实施和优化Microsoft解决方案。

咨询专家

获取微软技术最新资讯

AzureMicrosoft FoundryClaude Opus 4.6AI agentsgovernance

相关文章

Azure

Microsoft The Shift Podcast on Agentic AI Challenges

Microsoft has launched a new season of The Shift podcast focused on agentic AI, with eight weekly episodes exploring how AI agents use data, coordinate with each other, and depend on platforms like Postgres, Microsoft Fabric, and OneLake. The series matters because it highlights that deploying agents in enterprises is not just about models—it requires rethinking architecture, governance, security, and IT workflows across the full Azure and data stack.

Azure

Azure Agentic AI for Regulated Industry Modernization

Microsoft says Azure combined with agentic AI can help regulated industries modernize legacy systems faster by automating workload assessment, migration, and ongoing operations while maintaining compliance. The update matters because it positions cloud migration as more than a cost-saving exercise: for sectors like healthcare and other highly regulated industries, it is increasingly essential for resilience, governance, and readiness to deploy AI at scale.

Azure

Fireworks AI on Microsoft Foundry for Azure Inference

Microsoft has launched a public preview of Fireworks AI on Microsoft Foundry, bringing high-throughput, low-latency open-model inference to Azure through a single managed endpoint. It matters because enterprises can now access models like DeepSeek V3.2, gpt-oss-120b, Kimi K2.5, and MiniMax M2.5 with Azure’s governance, serverless or provisioned deployment options, and bring-your-own-weights support—making it easier to move open-model AI from experimentation into production.

Azure

Azure Copilot Migration Agent for App Modernization

Microsoft has introduced new public preview modernization agents in Azure Copilot and GitHub Copilot to help organizations automate migration and application transformation across discovery, assessment, planning, deployment, and code upgrades. The announcement matters because it aims to turn complex, fragmented modernization work into a coordinated AI-assisted workflow, helping enterprises move legacy infrastructure and applications to Azure faster and with clearer cost, dependency, and prioritization insights.

Azure

Azure IaaS Resource Center for Resilient Infrastructure

Microsoft has introduced the Azure IaaS Resource Center, a centralized hub for infrastructure teams to find design guidance, demos, architecture resources, and best practices for compute, storage, and networking. The launch matters because it reinforces Azure IaaS as a unified platform for building resilient, high-performance, and cost-optimized infrastructure, helping organizations better support everything from traditional business apps to AI workloads.

Azure

Microsoft Foundry ROI Study Shows 327% Enterprise AI Gains

A Forrester Total Economic Impact study commissioned around Microsoft Foundry found that a modeled enterprise could achieve 327% ROI over three years, break even in about six months, and realize $49.5 million in benefits from productivity and infrastructure savings. The results matter because they highlight how much enterprise AI costs are driven by developer time and fragmented tooling, suggesting that a unified platform like Foundry can help IT teams accelerate AI delivery while improving governance and efficiency.