Sign In
By signing in or creating an account, you agree with Associated Broadcasting Company's Terms & Conditions and Privacy Policy.
By signing in or creating an account, you agree with Associated Broadcasting Company's Terms & Conditions and Privacy Policy.
New Delhi: Microsoft, NVIDIA and Anthropic have announced one of the largest three-way AI partnerships seen so far, linking cloud, chips and model development at a scale that reflects how fast the AI industry is expanding. The companies revealed a series of long term deals that include billions of dollars in compute commitments, new chip collaborations and wider access to Anthropic’s Claude models on Microsoft’s cloud.
The announcement comes at a time when big tech firms are trying to secure huge amounts of compute power for training and running advanced AI models. With new competition building and investors starting to question the pace of AI spending, this partnership signals how the major players are trying to lock in supply chains early and reduce dependence on any single partner.
The biggest piece of the announcement is Anthropic agreeing to purchase 30 billion dollars worth of compute on Microsoft Azure, along with the option to contract up to one gigawatt of additional capacity. In simple terms, it is one of the largest AI compute commitments ever made. Microsoft said the Claude maker will scale its future models on Azure, powered by NVIDIA hardware.
Microsoft CEO Satya Nadella described the partnership by saying, “We’re increasingly going to be customers of each other. We will use Anthropic models, they will use our infrastructure and we’ll go to market together.”
Anthropic will also continue to work with Amazon as its primary cloud and training partner, which places the startup in a unique position with major deals across all three global cloud giants. At present, Claude will be the only frontier model available across Azure, AWS and Google Cloud.
For the first time, NVIDIA and Anthropic are entering a long term collaboration focused on designing future chips and optimising Claude models for better performance and efficiency. NVIDIA will invest up to 10 billion dollars in the AI firm, while providing Grace Blackwell and Vera Rubin supercomputing systems for Anthropic’s future workloads.
Industry executives estimate that one gigawatt of AI compute can cost between 20 billion and 25 billion dollars. This gives an idea of how massive the demand for chips has become as companies chase AGI-level performance.
The companies said the joint effort aims to improve total cost of ownership, boost speed and tailor future GPU architectures around Anthropic’s model needs.
Along with NVIDIA’s commitment, Microsoft will invest up to 5 billion dollars in Anthropic as part of the startup’s next funding round. Anthropic was recently valued at 183 billion dollars, and Reuters reported that the company expects to reach an annualised revenue run rate of around 26 billion dollars next year. It already serves over 300,000 enterprise and business customers.
The partnership also expands Microsoft’s Foundry programme, giving Azure clients access to Claude Opus 4.1, Claude Sonnet 4.5 and Claude Haiku 4.5. Microsoft said Claude will remain part of its Copilot family, including GitHub Copilot, Microsoft 365 Copilot and Copilot Studio.
The deal also arrives just weeks after OpenAI made major structural changes and signed a 38 billion dollar cloud agreement with Amazon in an effort to diversify beyond Microsoft. Analysts have warned that many recent arrangements in the AI world resemble “circular deals” that could inflate valuations and fuel bubble risk.
Still, companies like Anthropic are trying to stay ahead by spreading their dependencies. As Nadella said in his video message, OpenAI “remains a critical partner” even as Microsoft adds deeper ties with one of its biggest competitors.
