Microsoft Azure now offers Claude AI models via Anthropic partnership
▼ Summary
– Microsoft is partnering with Anthropic to bring its Claude AI models to Microsoft Foundry while Anthropic commits to purchasing $30 billion of Azure compute capacity.
– Anthropic will remain primarily partnered with Amazon as its cloud provider and training partner despite this new Microsoft collaboration.
– Nvidia is partnering with Anthropic to optimize models for future architectures, with Anthropic committing up to one gigawatt of compute using Nvidia Blackwell and Vera Rubin systems.
– Microsoft and Nvidia are investing $5 billion and up to $10 billion respectively in Anthropic as part of these partnerships.
– Microsoft’s new Anthropic deal follows recent OpenAI partnership changes that relaxed exclusivity terms and allowed Microsoft to pursue other AI collaborations.
A major new alliance between Microsoft and Anthropic is set to significantly expand the AI capabilities available on the Azure cloud platform. Microsoft Azure customers will now gain access to Anthropic’s advanced Claude AI models, including Claude Sonnet 4.5, Claude Opus 4.1, and Claude Haiku 4.5, through the Microsoft Foundry service. This strategic partnership is further cemented by Anthropic’s substantial commitment to utilize Azure’s infrastructure, planning to acquire $30 billion in Azure compute capacity and contract for additional capacity reaching up to one gigawatt.
The collaboration extends to include Nvidia, which is partnering with Anthropic to ensure the Claude models are finely tuned for optimal performance on upcoming Nvidia Blackwell and Vera Rubin architectures. To support this initiative, Nvidia is making a significant investment of up to $10 billion in Anthropic. Microsoft is also contributing a $5 billion investment, reinforcing its commitment to the partnership.
This new agreement with Anthropic arrives shortly after Microsoft solidified its long-term relationship with OpenAI. The revised deal with OpenAI grants Microsoft intellectual property rights to OpenAI’s models until 2032, a term that also covers any post-artificial general intelligence (AGI) models developed. A key aspect of this agreement was establishing a clear process for verifying the achievement of AGI, which will now be determined by an independent panel of experts.
Furthermore, the exclusivity clauses in the Microsoft-OpenAI partnership have been relaxed. This adjustment permits OpenAI to engage in collaborations with other third-party companies and to release open-weight models. Microsoft also relinquished its right of first refusal to serve as OpenAI’s primary compute provider, a change that has effectively opened the door for Microsoft to pursue its new strategic relationship with Anthropic.
Microsoft has already been integrating Anthropic’s technology more deeply into its ecosystem. Over recent months, the company has shown a growing preference for Claude models within its various Copilot services. For instance, the new auto AI model selector in Visual Studio Code has been observed favoring Anthropic’s Claude 4 over other models like GPT-5. Additionally, Microsoft has recently made Claude Sonnet 4 and Claude Opus 4.1 available to its Microsoft 365 Copilot user base, signaling a broader adoption of Anthropic’s AI solutions across its product lineup.
(Source: The Verge)
