Anthropic to Draw $15 Bn from Microsoft, Nvidia After Committing $30 Bn to Azure, 1GW of Nvidia Compute

Tech giants Microsoft and Nvidia are reportedly set to invest a combined $15 billion in Anthropic PBC. At the same time, Anthropic said it has committed to purchase roughly $30 billion worth of Azure compute capacity and to contract up to one gigawatt of Nvidia hardware in the coming years.

Under the partnership terms, Nvidia will invest up to $10 billion in the AI startup, while Microsoft will contribute up to $5 billion.

The move positions Anthropic closer to two of the strongest financial and technological backers of its biggest rival, OpenAI, marking a significant escalation in the AI industry’s race for compute power and strategic alliances.

Aim of Partnership

Microsoft and Anthropic in a blog stated the deal will broaden enterprise access to Claude by making the model family available through Microsoft Foundry and by expanding Claude-specific capabilities for Azure customers.

Microsoft also confirmed Claude will remain available inside its Copilot product family and that, with this move, Claude will be offered on all three leading cloud platforms, increasing model choice for enterprise users.

Nvidia-Anthropic Engineering Pact

Anthropic will run initial capacity on Nvidia’s Grace Blackwell and Vera Rubin systems, and the two firms will collaborate on co-design and optimisation so models and hardware better match each other for performance, efficiency and total cost of ownership.

The one-gigawatt figure refers to the scale of compute capacity Anthropic may contract as it scales Claude.

The arrangement is both strategic and financial. Beyond the compute purchase, Microsoft and Nvidia’s commitments, described by the companies as investments in Anthropic, bind key parts of the AI supply chain together, like cloud access, chips and model development.

Another Circular Deal?

The deal intensifies competition among entrenched AI ecosystems. Anthropic’s wider placement on Azure gives enterprises an alternative to models offered by other cloud–provider tie-ups, and Nvidia’s deeper engineering role signals vendors’ desire to tune hardware for specific, high-value workloads.

The pact also reflects the growing trend of “circular” AI economics, model makers buying cloud capacity and cloud/chip providers taking equity or making strategic investments to secure long-term demand.