Anthropic launched its Claude models in Microsoft Foundry on Azure on Monday, June 29, 2026, running the service on NVIDIA GB300 Blackwell Ultra GPU systems and turning a partnership first previewed in late 2025 into generally available enterprise infrastructure for Azure-native AI applications. The announcement is less a simple model-catalog update than a statement about where frontier AI is being industrialized. Claude is no longer merely a rival chatbot or an API endpoint; it is becoming another managed workload inside the Microsoft cloud stack.
That matters because Azure has spent years being identified with OpenAI, sometimes to the point where Microsoft’s AI strategy looked like a single-vendor bet wrapped in a hyperscale cloud. The arrival of Claude on GB300-backed Azure infrastructure changes that framing. Microsoft is not abandoning OpenAI, but it is making clear that the next phase of enterprise AI will be fought on choice, governance, procurement, and compute supply — not just model benchmarks.
The headline version is straightforward: Azure customers can now use Anthropic’s Claude family through Microsoft Foundry, with the models hosted on Azure and operated through Microsoft’s enterprise AI platform. The more interesting version is that Microsoft is turning model choice into a native cloud primitive.
For years, CIOs have complained that AI adoption is less about finding a clever demo than getting legal, security, billing, identity, and observability aligned. A developer can sign up for a model API in minutes; an enterprise may need months to approve the same thing. By putting Claude inside Microsoft Foundry, Microsoft is trying to collapse that distance.
The practical pitch is familiar to anyone who has worked inside a Microsoft-heavy organization. If the model is available through Azure agreements, tied into Microsoft identity, and governed through the same cloud controls as other workloads, it becomes easier for IT to say yes. That does not make Claude risk-free, cheap, or magically compliant, but it moves the conversation from “new vendor exception” to “approved platform capability.”
This is where Microsoft’s advantage is most visible. Azure is not just selling access to tokens. It is selling access to tokens that fit into existing enterprise machinery: procurement, cost management, role-based access, audit trails, private networking patterns, and the growing governance layer around AI agents.
Amazon remains deeply tied to Anthropic’s infrastructure strategy, and Google Cloud has also been an important channel for Claude. That makes Anthropic one of the rare frontier AI companies trying to be genuinely multi-cloud at the top end of the market. The Azure launch extends that posture into the Microsoft enterprise base, where many large customers are already standardizing AI projects around Foundry and Copilot-related tooling.
The upside is obvious: Claude can meet customers where they already are. The risk is that Anthropic becomes dependent on a set of infrastructure giants whose incentives overlap with, but do not perfectly match, its own. Microsoft, Amazon, Google, and NVIDIA all want frontier AI demand to grow. They also all want that demand to reinforce their own platforms, chips, clouds, and developer ecosystems.
Anthropic’s challenge is to benefit from the reach of the hyperscalers without becoming a feature inside someone else’s stack. That is a delicate balance. The Azure launch gives Claude a bigger enterprise lane, but it also embeds the model more deeply into the competitive politics of cloud computing.
GB300 Blackwell Ultra systems are designed for the kind of inference and reasoning-heavy workloads that modern AI agents increasingly demand. These are not just chat completions. Enterprise AI systems are chaining tool calls, reading large document sets, generating and testing code, interacting with business systems, and sometimes running for long stretches before returning a result.
That shift changes the compute equation. A model that performs well in a benchmark can still feel sluggish, expensive, or unreliable if the serving infrastructure is constrained. For customers building AI into support desks, developer workflows, security operations, financial analysis, or compliance review, infrastructure performance is not an abstraction. It shows up as latency, throughput, quota limits, and invoices.
Microsoft’s message is that Claude on Azure is not merely available; it is available on current-generation NVIDIA infrastructure built for high-demand AI workloads. NVIDIA’s message is equally clear: even model companies that previously emphasized other hardware strategies are finding their way onto NVIDIA systems when the market demands scale, performance, and developer familiarity.
That does not diminish OpenAI’s importance. GPT models still sit at the center of many Microsoft AI experiences, and Microsoft has every reason to keep that relationship strong. But enterprise customers rarely want a theology of model selection. They want leverage, optionality, and fallback paths.
Claude’s presence in Foundry gives Microsoft a more credible answer to customers who do not want all AI roads to lead to one vendor. Some workloads may favor Claude. Others may favor OpenAI, smaller open models, domain-specific models, or specialized local deployments. The strategic prize for Microsoft is not winning every model comparison; it is making Azure the place where those comparisons happen.
This is a subtle but important shift. If Azure becomes the neutral-ish enterprise layer where organizations test, deploy, monitor, and swap models, Microsoft can profit even when the winning model is not one it created. That is the cloud platform dream updated for the AI era.
Claude Code and similar agentic coding tools have been popular partly because they meet developers in the terminal and editor rather than forcing every task through a browser. But unmanaged AI coding tools create predictable headaches for IT: personal API keys, unclear billing, weak auditability, and little control over which source code or internal data leaves the organization.
Putting Claude behind Microsoft Foundry and Azure API Management offers a cleaner pattern. A company can route developer access through Entra authentication, apply rate limits, track token usage, centralize billing, and avoid scattering secrets across laptops. That does not solve every data governance problem, but it gives administrators familiar controls instead of asking them to bless a shadow AI workflow.
This is where the Azure launch may become more meaningful than the model announcement itself. Developers care whether the tool works. Security teams care whether the tool can be governed. Finance teams care whether usage can be measured. Microsoft is trying to make those interests compatible enough that AI coding agents can move from tolerated experiments to sanctioned infrastructure.
This is why the governance layer matters. Once models can call tools, write code, query databases, and interact with enterprise systems, access control becomes as important as intelligence. A powerful model with weak boundaries is not an assistant; it is an unmanaged automation surface.
Microsoft knows this terrain. Its enterprise business is built on the premise that organizations will pay for control. Entra, Purview, Defender, Intune, Azure Policy, and the rest of the Microsoft management universe all exist because large customers do not simply buy capabilities; they buy ways to constrain capabilities.
Claude’s arrival in Foundry fits that pattern. The model is the visible feature, but the platform story is about containing AI inside systems that administrators can understand. That will be especially appealing to organizations that like Claude’s capabilities but have hesitated to adopt another standalone AI vendor relationship.
Enterprises will need to watch how quickly agentic usage changes their cost profile. A human asking a model for a summary is one kind of workload. A coding agent iterating through a repository, running tests, reading logs, and generating patches is another. A research agent that performs multi-step retrieval and synthesis across internal systems is another still.
Microsoft’s cloud billing model can help organizations track and allocate those costs, but it will not eliminate them. The next phase of AI governance will be as much about budget controls as safety controls. Token quotas, per-user limits, model routing, caching, and workload-specific model selection will become practical necessities rather than optimization hobbies.
This is also where smaller models and mixed-model architectures remain important. The presence of Claude Opus or Sonnet-class models in Foundry does not mean every task should use the largest available model. Sensible enterprises will route routine work to cheaper models, reserve frontier systems for high-value tasks, and continuously test whether the performance premium is justified.
Anthropic has worked with multiple hardware strategies, including custom accelerators through major cloud partners. Yet this launch places Claude on NVIDIA GB300 systems inside Azure. That does not mean NVIDIA owns every future workload, but it shows how difficult it is to avoid NVIDIA entirely at the frontier.
The reason is not just raw silicon. NVIDIA’s advantage includes networking, system design, libraries, developer tooling, and operational familiarity at scale. When hyperscalers are trying to stand up massive AI clusters and serve demanding customers, the safest answer is often the one with the deepest ecosystem.
For Microsoft, NVIDIA infrastructure gives Azure a credible performance story. For Anthropic, it offers another path to capacity. For enterprise customers, it provides a reassuringly mainstream foundation for workloads that may become business-critical. The whole arrangement demonstrates why NVIDIA continues to sit in the middle of AI’s cloud economy, even when the branding belongs to someone else.
The result is not a clean stack; it is a web of commercial interdependence. Rivals are customers. Suppliers are investors. Cloud platforms host model companies that compete with their own AI products. Every major player is trying to avoid being locked out of the next layer of value.
For enterprise buyers, this messiness can be useful. Competition among model providers may improve pricing, capabilities, and availability. Multi-cloud availability reduces the risk that a single vendor relationship dictates every AI architecture decision. But it also makes due diligence harder.
Customers will need to understand not just which model performs best, but where it runs, who operates it, how data is handled, which regions are available, what compliance commitments apply, and how outages or policy changes propagate through the stack. The AI procurement checklist is becoming longer, not shorter.
Foundry is aiming for the same role in AI. The platform does not pretend that enterprises will standardize on a single model, a single agent framework, or a single data pattern. Instead, it offers a place to manage the sprawl.
Claude on Azure strengthens that pitch. Microsoft can now tell customers that they do not have to choose between the OpenAI ecosystem and Anthropic’s model family at the platform level. They can evaluate both inside a Microsoft-governed environment and let workloads determine the winner.
That is a powerful argument, especially for organizations already committed to Azure. The danger is that the platform itself becomes another layer of lock-in. Model choice inside a single cloud is still cloud dependency. For some customers, that trade-off will be acceptable. For others, especially those with strict portability or regulatory concerns, the architecture will need more scrutiny.
But the presence of a trusted cloud platform does not automatically make AI safe. Models can still produce flawed code, mishandle ambiguous instructions, expose sensitive information through poorly designed workflows, or take unsafe actions when connected to tools. The security boundary is not the model card; it is the whole system around the model.
That means organizations need to test Claude-powered workflows like they would test any automation touching production systems. They need permissions scoped tightly, logs retained appropriately, prompts and tool outputs inspected, and human approval inserted where mistakes would be costly. AI agents should earn autonomy gradually, not receive it by default because the demo was impressive.
The best use of Microsoft’s governance stack is not to rubber-stamp AI adoption. It is to make experimentation observable, constrained, and reversible. That is how enterprises can learn where Claude is genuinely useful without turning every department into its own unsupervised AI lab.
Microsoft, Anthropic, and NVIDIA are each selling a different version of the same future: AI models as utility-scale services, running on specialized hardware, governed by enterprise platforms, and embedded into the daily work of developers, analysts, administrators, and knowledge workers. The winners will not be decided by launch-day claims alone. They will be decided by whether these systems can deliver reliable value under the boring but unforgiving conditions of real IT: budgets, audits, outages, permissions, latency, compliance, and users who expect the magic to work every morning.
That matters because Azure has spent years being identified with OpenAI, sometimes to the point where Microsoft’s AI strategy looked like a single-vendor bet wrapped in a hyperscale cloud. The arrival of Claude on GB300-backed Azure infrastructure changes that framing. Microsoft is not abandoning OpenAI, but it is making clear that the next phase of enterprise AI will be fought on choice, governance, procurement, and compute supply — not just model benchmarks.
Microsoft Turns Model Choice Into a Cloud Feature
The headline version is straightforward: Azure customers can now use Anthropic’s Claude family through Microsoft Foundry, with the models hosted on Azure and operated through Microsoft’s enterprise AI platform. The more interesting version is that Microsoft is turning model choice into a native cloud primitive.For years, CIOs have complained that AI adoption is less about finding a clever demo than getting legal, security, billing, identity, and observability aligned. A developer can sign up for a model API in minutes; an enterprise may need months to approve the same thing. By putting Claude inside Microsoft Foundry, Microsoft is trying to collapse that distance.
The practical pitch is familiar to anyone who has worked inside a Microsoft-heavy organization. If the model is available through Azure agreements, tied into Microsoft identity, and governed through the same cloud controls as other workloads, it becomes easier for IT to say yes. That does not make Claude risk-free, cheap, or magically compliant, but it moves the conversation from “new vendor exception” to “approved platform capability.”
This is where Microsoft’s advantage is most visible. Azure is not just selling access to tokens. It is selling access to tokens that fit into existing enterprise machinery: procurement, cost management, role-based access, audit trails, private networking patterns, and the growing governance layer around AI agents.
Anthropic Gets the Enterprise Door Without Giving Up the Rest of the House
For Anthropic, Azure availability is a distribution win with strategic complications. Claude already had strong enterprise credibility, especially among developers and organizations that value long-context reasoning, coding, and agentic workflows. But the company’s cloud posture has always been more complicated than a simple “pick one hyperscaler” story.Amazon remains deeply tied to Anthropic’s infrastructure strategy, and Google Cloud has also been an important channel for Claude. That makes Anthropic one of the rare frontier AI companies trying to be genuinely multi-cloud at the top end of the market. The Azure launch extends that posture into the Microsoft enterprise base, where many large customers are already standardizing AI projects around Foundry and Copilot-related tooling.
The upside is obvious: Claude can meet customers where they already are. The risk is that Anthropic becomes dependent on a set of infrastructure giants whose incentives overlap with, but do not perfectly match, its own. Microsoft, Amazon, Google, and NVIDIA all want frontier AI demand to grow. They also all want that demand to reinforce their own platforms, chips, clouds, and developer ecosystems.
Anthropic’s challenge is to benefit from the reach of the hyperscalers without becoming a feature inside someone else’s stack. That is a delicate balance. The Azure launch gives Claude a bigger enterprise lane, but it also embeds the model more deeply into the competitive politics of cloud computing.
GB300 Is the Quiet Center of the Announcement
The NVIDIA GB300 detail is not decorative. In 2026, the story of frontier AI is inseparable from the story of accelerated infrastructure, and Microsoft wants customers to see Azure as a place where the newest model workloads can run at production scale.GB300 Blackwell Ultra systems are designed for the kind of inference and reasoning-heavy workloads that modern AI agents increasingly demand. These are not just chat completions. Enterprise AI systems are chaining tool calls, reading large document sets, generating and testing code, interacting with business systems, and sometimes running for long stretches before returning a result.
That shift changes the compute equation. A model that performs well in a benchmark can still feel sluggish, expensive, or unreliable if the serving infrastructure is constrained. For customers building AI into support desks, developer workflows, security operations, financial analysis, or compliance review, infrastructure performance is not an abstraction. It shows up as latency, throughput, quota limits, and invoices.
Microsoft’s message is that Claude on Azure is not merely available; it is available on current-generation NVIDIA infrastructure built for high-demand AI workloads. NVIDIA’s message is equally clear: even model companies that previously emphasized other hardware strategies are finding their way onto NVIDIA systems when the market demands scale, performance, and developer familiarity.
The OpenAI Era Gives Way to the Portfolio Era
Microsoft’s partnership with OpenAI remains central to its AI identity, but the company’s platform strategy has been broadening. Foundry is the clearest expression of that shift. It is less about one model family and more about giving enterprises a control plane for models, tools, agents, data, and governance.That does not diminish OpenAI’s importance. GPT models still sit at the center of many Microsoft AI experiences, and Microsoft has every reason to keep that relationship strong. But enterprise customers rarely want a theology of model selection. They want leverage, optionality, and fallback paths.
Claude’s presence in Foundry gives Microsoft a more credible answer to customers who do not want all AI roads to lead to one vendor. Some workloads may favor Claude. Others may favor OpenAI, smaller open models, domain-specific models, or specialized local deployments. The strategic prize for Microsoft is not winning every model comparison; it is making Azure the place where those comparisons happen.
This is a subtle but important shift. If Azure becomes the neutral-ish enterprise layer where organizations test, deploy, monitor, and swap models, Microsoft can profit even when the winning model is not one it created. That is the cloud platform dream updated for the AI era.
Windows Developers Will Feel This Through Tools, Not Press Releases
For WindowsForum readers, the most immediate impact will not be a data-center spec sheet. It will be the way Claude becomes easier to route into developer workflows that already live on Windows, Visual Studio Code, GitHub, Azure DevOps, Microsoft Entra ID, and corporate networks.Claude Code and similar agentic coding tools have been popular partly because they meet developers in the terminal and editor rather than forcing every task through a browser. But unmanaged AI coding tools create predictable headaches for IT: personal API keys, unclear billing, weak auditability, and little control over which source code or internal data leaves the organization.
Putting Claude behind Microsoft Foundry and Azure API Management offers a cleaner pattern. A company can route developer access through Entra authentication, apply rate limits, track token usage, centralize billing, and avoid scattering secrets across laptops. That does not solve every data governance problem, but it gives administrators familiar controls instead of asking them to bless a shadow AI workflow.
This is where the Azure launch may become more meaningful than the model announcement itself. Developers care whether the tool works. Security teams care whether the tool can be governed. Finance teams care whether usage can be measured. Microsoft is trying to make those interests compatible enough that AI coding agents can move from tolerated experiments to sanctioned infrastructure.
The Agent Boom Makes Governance the Product
The industry’s language has shifted from chatbots to agents because the economic promise has shifted from answering questions to doing work. An agent that can investigate an incident, refactor a codebase, draft a migration plan, or reconcile a spreadsheet has more business value than a chatbot that summarizes a memo. It also has more capacity to make mistakes at scale.This is why the governance layer matters. Once models can call tools, write code, query databases, and interact with enterprise systems, access control becomes as important as intelligence. A powerful model with weak boundaries is not an assistant; it is an unmanaged automation surface.
Microsoft knows this terrain. Its enterprise business is built on the premise that organizations will pay for control. Entra, Purview, Defender, Intune, Azure Policy, and the rest of the Microsoft management universe all exist because large customers do not simply buy capabilities; they buy ways to constrain capabilities.
Claude’s arrival in Foundry fits that pattern. The model is the visible feature, but the platform story is about containing AI inside systems that administrators can understand. That will be especially appealing to organizations that like Claude’s capabilities but have hesitated to adopt another standalone AI vendor relationship.
The Cost Story Is Still Unwritten
The hard part is economics. Frontier AI is expensive to train, expensive to serve, and expensive to scale globally. GB300 infrastructure improves performance, but it does not make high-end AI free. If anything, the more capable the models become, the more ambitious and compute-hungry the workloads tend to get.Enterprises will need to watch how quickly agentic usage changes their cost profile. A human asking a model for a summary is one kind of workload. A coding agent iterating through a repository, running tests, reading logs, and generating patches is another. A research agent that performs multi-step retrieval and synthesis across internal systems is another still.
Microsoft’s cloud billing model can help organizations track and allocate those costs, but it will not eliminate them. The next phase of AI governance will be as much about budget controls as safety controls. Token quotas, per-user limits, model routing, caching, and workload-specific model selection will become practical necessities rather than optimization hobbies.
This is also where smaller models and mixed-model architectures remain important. The presence of Claude Opus or Sonnet-class models in Foundry does not mean every task should use the largest available model. Sensible enterprises will route routine work to cheaper models, reserve frontier systems for high-value tasks, and continuously test whether the performance premium is justified.
NVIDIA Wins Even When the Cloud Logos Change
NVIDIA’s role in the announcement is another reminder that the AI market’s most durable choke point is still compute. Cloud providers compete fiercely with one another, and model providers compete even more visibly, but many of the biggest roads continue to pass through NVIDIA’s hardware and software ecosystem.Anthropic has worked with multiple hardware strategies, including custom accelerators through major cloud partners. Yet this launch places Claude on NVIDIA GB300 systems inside Azure. That does not mean NVIDIA owns every future workload, but it shows how difficult it is to avoid NVIDIA entirely at the frontier.
The reason is not just raw silicon. NVIDIA’s advantage includes networking, system design, libraries, developer tooling, and operational familiarity at scale. When hyperscalers are trying to stand up massive AI clusters and serve demanding customers, the safest answer is often the one with the deepest ecosystem.
For Microsoft, NVIDIA infrastructure gives Azure a credible performance story. For Anthropic, it offers another path to capacity. For enterprise customers, it provides a reassuringly mainstream foundation for workloads that may become business-critical. The whole arrangement demonstrates why NVIDIA continues to sit in the middle of AI’s cloud economy, even when the branding belongs to someone else.
The Competitive Map Gets Messier, Not Cleaner
This launch also complicates the tidy narratives people like to tell about AI alliances. Microsoft backs OpenAI but now offers Claude more deeply in Azure. Amazon remains a major Anthropic partner while Microsoft sells Anthropic access to its customers. Google competes with both Microsoft and Amazon while also distributing Claude through Vertex AI. NVIDIA supplies the hardware layer while investing across the ecosystem.The result is not a clean stack; it is a web of commercial interdependence. Rivals are customers. Suppliers are investors. Cloud platforms host model companies that compete with their own AI products. Every major player is trying to avoid being locked out of the next layer of value.
For enterprise buyers, this messiness can be useful. Competition among model providers may improve pricing, capabilities, and availability. Multi-cloud availability reduces the risk that a single vendor relationship dictates every AI architecture decision. But it also makes due diligence harder.
Customers will need to understand not just which model performs best, but where it runs, who operates it, how data is handled, which regions are available, what compliance commitments apply, and how outages or policy changes propagate through the stack. The AI procurement checklist is becoming longer, not shorter.
Azure’s Real Bet Is That Enterprises Prefer Managed Complexity
Microsoft has rarely won by making technology simple in the consumer sense. It wins by making complexity manageable for institutions. Windows, Office, Active Directory, Exchange, SharePoint, Azure, and Microsoft 365 all became enterprise defaults because they matched the messy reality of large organizations.Foundry is aiming for the same role in AI. The platform does not pretend that enterprises will standardize on a single model, a single agent framework, or a single data pattern. Instead, it offers a place to manage the sprawl.
Claude on Azure strengthens that pitch. Microsoft can now tell customers that they do not have to choose between the OpenAI ecosystem and Anthropic’s model family at the platform level. They can evaluate both inside a Microsoft-governed environment and let workloads determine the winner.
That is a powerful argument, especially for organizations already committed to Azure. The danger is that the platform itself becomes another layer of lock-in. Model choice inside a single cloud is still cloud dependency. For some customers, that trade-off will be acceptable. For others, especially those with strict portability or regulatory concerns, the architecture will need more scrutiny.
Security Teams Should Welcome the Control and Distrust the Hype
Security-minded readers should treat this announcement with cautious optimism. The ability to bring Claude into an Azure-governed environment is useful. It can reduce shadow AI usage, improve access control, and make monitoring more realistic.But the presence of a trusted cloud platform does not automatically make AI safe. Models can still produce flawed code, mishandle ambiguous instructions, expose sensitive information through poorly designed workflows, or take unsafe actions when connected to tools. The security boundary is not the model card; it is the whole system around the model.
That means organizations need to test Claude-powered workflows like they would test any automation touching production systems. They need permissions scoped tightly, logs retained appropriately, prompts and tool outputs inspected, and human approval inserted where mistakes would be costly. AI agents should earn autonomy gradually, not receive it by default because the demo was impressive.
The best use of Microsoft’s governance stack is not to rubber-stamp AI adoption. It is to make experimentation observable, constrained, and reversible. That is how enterprises can learn where Claude is genuinely useful without turning every department into its own unsupervised AI lab.
The Claude-on-Azure Deal Leaves IT With Fewer Excuses and More Decisions
The launch narrows the gap between AI ambition and enterprise deployment reality. Claude is now easier for Azure customers to procure, govern, and route into production-style workflows, which means the harder questions move from availability to architecture.- Organizations already standardized on Azure can evaluate Claude without building a separate vendor and billing path from scratch.
- Developers gain a more enterprise-friendly route to Claude-powered coding and agent workflows through Microsoft’s platform controls.
- Administrators should treat cost governance, identity enforcement, logging, and rate limits as first-order design requirements.
- Microsoft strengthens Azure by making it a portfolio platform for frontier models rather than a single-model showroom.
- Anthropic gains distribution, but its growing dependence on hyperscale infrastructure partners will remain a strategic tension.
- NVIDIA’s GB300 role reinforces that frontier AI competition still depends heavily on access to the newest accelerated compute.
Microsoft, Anthropic, and NVIDIA are each selling a different version of the same future: AI models as utility-scale services, running on specialized hardware, governed by enterprise platforms, and embedded into the daily work of developers, analysts, administrators, and knowledge workers. The winners will not be decided by launch-day claims alone. They will be decided by whether these systems can deliver reliable value under the boring but unforgiving conditions of real IT: budgets, audits, outages, permissions, latency, compliance, and users who expect the magic to work every morning.
References
- Primary source: investing.com
Published: Mon, 29 Jun 2026 17:50:07 GMT
Loading…
www.investing.com - Related coverage: tomshardware.com
- Related coverage: nvidia.com
NVIDIA GB300 NVL72
The NVIDIA GB300 NVL72 features a fully liquid-cooled, rack-scale design that unifies 72 NVIDIA Blackwell Ultra GPUs and 36 Arm based NVIDIA Grace CPUs in a single platform.www.nvidia.com - Related coverage: blogs.nvidia.com
Loading…
blogs.nvidia.com - Related coverage: id.investing.com
Loading…
id.investing.com - Related coverage: gadgets360.com
Loading…
www.gadgets360.com
- Official source: techcommunity.microsoft.com
Loading…
techcommunity.microsoft.com - Related coverage: cryptobriefing.com
Anthropic's Claude models now run on Nvidia GB300 Blackwell Ultra systems via Microsoft Azure
Anthropic's Claude AI models now run on Nvidia GB300 Blackwell Ultra via Azure, backed by $15B in investments and a $30B compute commitment reshaping AIcryptobriefing.com - Official source: news.microsoft.com
Loading…
news.microsoft.com - Related coverage: computerbase.de
Loading…
www.computerbase.de - Related coverage: techradar.com
Anthropic locks in massive Azure deal to fuel Claude expansion across global clouds and reshape enterprise AI access worldwide | TechRadar
Claude models integrate into the Microsoft Foundry platform for enterprise deploymentwww.techradar.com - Related coverage: axios.com
Anthropic lands $15 billion investment from Microsoft, Nvidia
The move is the latest in a series of deals that have all the big players partnering with one another.www.axios.com
- Related coverage: windowscentral.com
NVIDIA joins Microsoft’s push on Claude — piling billions into Anthropic’s future | Windows Central
Claude’s arrival on Azure signals a major shift in the competitive AI cloud landscape.www.windowscentral.com - Related coverage: docs.nvidia.com
- Related coverage: academy.nvidia.com
AI Infrastructure Training for Professionals
Master enterprise AI infrastructure. Learn to deploy scalable clusters, manage GPU orchestration, and optimize production workload performance with NVIDIA.academy.nvidia.com - Related coverage: arturmarkus.com
- Related coverage: nvidianews.nvidia.com
- Official source: anthropic.com
Claude now available in Microsoft Foundry and Microsoft 365 Copilot
Claude Sonnet 4.5, Haiku 4.5, and Opus 4.1 models are now available in public preview in Microsoft Foundry, where Azure customers can build production applications and enterprise agents.
www.anthropic.com
- Official source: azure.microsoft.com
Loading…
azure.microsoft.com - Official source: www-cdn.anthropic.com
Loading…
www-cdn.anthropic.com