惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

D
DataBreaches.Net
GbyAI
GbyAI
aimingoo的专栏
aimingoo的专栏
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
月光博客
月光博客
大猫的无限游戏
大猫的无限游戏
M
MIT News - Artificial intelligence
腾讯CDC
博客园 - Franky
Engineering at Meta
Engineering at Meta
C
Check Point Blog
T
The Blog of Author Tim Ferriss
有赞技术团队
有赞技术团队
Microsoft Azure Blog
Microsoft Azure Blog
MyScale Blog
MyScale Blog
I
InfoQ
Blog — PlanetScale
Blog — PlanetScale
P
Proofpoint News Feed
The GitHub Blog
The GitHub Blog
N
Netflix TechBlog - Medium
Last Week in AI
Last Week in AI
S
SegmentFault 最新的问题
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
WordPress大学
WordPress大学

TechWire Asia

Microsoft to deploy AMD Helios AI systems on Azure Microsoft ERP Implementation Partners Roundup: Which Is Best For Your Business? Semiconductor talent retention now comes with a price tag, and it keeps climbing AI token usage is not productivity. Here's what to measure instead Oracle brings governed AI agents to Fusion Applications Nvidia expands Japan AI infrastructure and robotics push AI Appreciation Day 2026 puts trust and governance in focus NVIDIA pours its full stack into Japan. The flip side of its China lockout? Malaysia's digital regulations are becoming a real cost for its startups Malaysia's AI data center vision: How EdgeConneX is building for the future Southeast Asia tech funding doubled to $7.4 billion. One company took most of it SK Hynix's Nasdaq listing raises $26.5 billion to fund Korea's AI memory expansion OpenAI launches GPT-5.6 for coding, cyber and science Meta rolls out Muse Image AI model for Instagram, WhatsApp, and advertisers Malaysia businesses face AI and password cybersecurity risks How AI workloads will test APAC mobile networks Enterprise AI costs don't have to spiral, argues ManageEngine Microsoft launches $2.5B Frontier Company for enterprise AI FIFA World Cup: How To Win Fans in APAC With Technology Kanga enters a new phase of global growth and launches Kanga Global Vertiv ramps up manufacturing in Johor's tightening data centre market U Mobile completes migration to own ULTRA5G network after DNB exit Asia built the AI infrastructure boom. The BIS just flagged who's exposed if it stalls. Why Apple is lobbying Washington to buy China’s memory chips Nvidia-backed Firmus plans 170,000-GPU Batam AI data centre Taiwan robot makers march into humanoid systems IBM claims world’s first sub-1 nm chip technology using nanostack design Can Alibaba bridge Malaysia’s SME talent gap via agentic AI for business? Huawei’s new tech explains why mobile AI network tech is no longer optional Apple-Intel chip deal faces years-long production timeline
Anthropic Claude models launch in Microsoft Foundry on Azure
Muhammad Zulhusni · 2026-06-30 · via TechWire Asia
  • Anthropic Claude is now available in Microsoft Foundry with Azure controls.
  • The rollout includes NVIDIA GB300 infrastructure and Foundry Agent Service support.

Anthropic’s Claude models are now generally available in Microsoft Foundry, with the service hosted on Azure. Organisations can run Claude within their Azure environment using existing Microsoft authentication, billing, and governance controls.

The rollout adds Claude to Microsoft Foundry’s broader model catalogue.

According to NVIDIA, Claude models in Microsoft Foundry are hosted on Azure infrastructure running NVIDIA GB300 Blackwell Ultra GPUs. The deployment uses NVIDIA GB300 NVL72 systems and NVIDIA Quantum-X800 InfiniBand networking.

The launch includes Claude Opus 4.8 and Claude Haiku 4.5 through the Messages API. The models support prompt caching, extended thinking, and tool streaming.

Azure controls and data options

Claude in Microsoft Foundry works with Azure identity, networking, and governance systems. Customers can authenticate with Microsoft Entra ID, apply Azure role-based access controls, and manage access through existing governance policies.

Customers receive billing through Azure, while eligible organisations with a Microsoft Enterprise Agreement can apply Claude usage against their Azure commitment. Claude usage is billed through Claude Consumption Units, with charges appearing on the customer’s Azure bill.

Microsoft said MACC drawdown and per-model usage detail in Foundry remain unchanged.

Customers can choose between Global and US data zones for inference processing, including workloads with data residency requirements.

In the Azure-hosted option, Azure infrastructure processes prompts and outputs, including request ingress, API services, and GPU inference. In the Anthropic-hosted option, Anthropic infrastructure processes prompts and outputs, and data may be processed outside Azure.

Microsoft’s documentation says Anthropic remains the seller and operator of Claude models in Microsoft Foundry and acts as an independent data processor for prompts and outputs.

Microsoft said zero data retention is available for high-sensitivity workloads. This means prompts and completions are not retained by Anthropic after the API call is completed.

Agent workflows and infrastructure

Microsoft said Foundry Agent Service can use Claude as the reasoning core for multi-step planning, tool use, and task execution across enterprise systems.

Microsoft also said Foundry includes a model router that can direct queries to the most appropriate Claude model. The company said the feature can reduce costs by up to 50%.

NVIDIA said enterprises can use its Secure Agent Workspace Reference Design to run Claude agents on Azure in governed environments. The reference design covers infrastructure-level controls for identity, network access, credentials, and runtime policy.

Hosting routes and customer use

Microsoft Foundry offers two Claude hosting options. The Azure-hosted route runs processing on Azure infrastructure with Microsoft authentication, billing, governance, and data zone controls.

The Anthropic-hosted route, previously available through the Foundry Preview, remains available for Anthropic API features or Claude models not yet supported in the Azure-hosted service. Anthropic said it aims to bring feature and model parity between the Azure-hosted and Anthropic-hosted options over time.

Momentic said it uses Claude Opus models through Microsoft Foundry to support software testing workflows and now serves millions of tokens per minute. The company said its customers describe tests in plain English, while Momentic runs through the interface to verify software before release.

Everstar said it used Anthropic and Azure for nuclear safety analysis work, according to Microsoft’s announcement.

Partnership context

The Azure deployment follows a November partnership between Microsoft, NVIDIA, and Anthropic covering Claude availability on NVIDIA accelerated computing. Reuters reported that Microsoft and NVIDIA planned to invest up to US$15 billion in Anthropic under the partnership, while Anthropic committed to buy US$30 billion in Azure compute capacity.

Anthropic said it plans to expand Claude’s availability in Microsoft Foundry over time.

Want to experience the full spectrum of enterprise technology innovation? Join TechEx in Amsterdam, California, and London. Covering AI, Big Data, Cyber Security, IoT, Digital Transformation, Intelligent Automation, Edge Computing, and Data Centres, TechEx brings together global leaders to share real-world use cases and in-depth insights. Click here for more information.

Tech Wire Asia is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.