Microsoft, NVIDIA and Anthropic have unveiled a landmark set of partnerships that significantly reshape how Claude’s frontier AI models will be scaled, deployed and optimised across cloud platforms. At the core of this expanded alliance is Anthropic’s major commitment to run its next wave of Claude systems on Microsoft Azure — a move enabled by NVIDIA’s latest infrastructure stack, including Grace Blackwell and Vera Rubin supercomputing technologies.
As part of the agreement, Anthropic has committed $30 billion in Azure compute with an option to secure up to one gigawatt of additional capacity. “We’re working to broaden access to Claude for organisations building with AI,” Anthropic CEO Dario Amodei said, speaking alongside Satya Nadella and Jensen Huang during the joint announcement.
Microsoft and NVIDIA are also preparing multi-billion-dollar investments into Anthropic, following the company’s $183 billion post-money valuation achieved during its recent Series F fundraise. The collaboration goes deeper still — marking the first time Anthropic and NVIDIA will co-design and engineer AI technology together, aligning Claude model development with NVIDIA’s chip architectures while NVIDIA optimises future processors for Anthropic’s training demands. Jensen Huang noted that this expanded partnership is designed to meet surging global compute needs and usher in the next generation of AI systems.
Microsoft is further broadening enterprise access to Claude’s top-tier models — including Sonnet 4.5, Opus 4.1 and Haiku 4.5 — via Microsoft Foundry, with Satya Nadella emphasising that “we want developers and enterprises to have choice in the models they use.”
Even as Amazon remains Anthropic’s primary cloud and training partner, the company is rapidly advancing its $50 billion U.S. compute infrastructure expansion across Texas and New York. These specialised data centres are expected to generate thousands of jobs while reinforcing national AI capacity under the federal government’s broader AI Action Plan.
