Microsoft has announced new strategic partnerships with Anthropic and OpenAI, further expanding its AI ecosystem. In a significant move, Anthropic has committed to acquiring $30 billion of Azure compute capacity and potentially up to one gigawatt in additional compute power. This collaboration enables Microsoft Foundry clients to utilize Anthropic’s advanced Claude models, such as Claude Sonnet 4.5, Claude Opus 4.1, and Claude Haiku 4.5, directly within Microsoft’s AI infrastructure.
While Anthropic will continue to leverage Amazon as its primary cloud provider and training partner, Nvidia has joined forces with Anthropic to enhance the optimization of Anthropic’s models for future Nvidia architectures. Furthermore, Anthropic has pledged to utilize Nvidia Blackwell and Vera Rubin systems for compute capacity, with Nvidia injecting up to $10 billion in investments and Microsoft contributing $5 billion to Anthropic.
This collaboration follows closely on the heels of Microsoft’s updated partnership with OpenAI, which recently underwent a for-profit restructuring and extended its agreement with Microsoft. The enhanced deal with OpenAI grants Microsoft intellectual property rights to models from OpenAI until 2032, encompassing post-AGI models. Notably, Microsoft and OpenAI have adjusted their partnership terms to allow OpenAI more flexibility in collaborating with external parties and releasing open weight models, relinquishing Microsoft’s previous right of first refusal to provide compute services to OpenAI.
Source: The Verge