Anthropic is in early discussions with Google on a cloud-computing agreement valued in the high tens of billions of dollars, according to a Tuesday (Oct. 21) Bloomberg report. The deal would give the artificial intelligence company access to Google’s tensor processing units, custom chips designed to accelerate machine learning, deepening a partnership that already includes billions in investments. Talks are preliminary but highlight how access to proprietary compute infrastructure has become a defining advantage in the AI race.
Google has invested about $3 billion in Anthropic to date and remains one of its key cloud providers. A larger deal could extend Google’s reach in the generative AI infrastructure market, where computing capacity and chip availability now determine who can train and deploy the most advanced models. The potential partnership comes as cloud giants and AI developers increasingly lock in long-term supply agreements to manage demand for high-performance hardware.
Anthropic’s rapid ascent reflects the growing economics of AI scale. The company recently raised $13 billion, bringing its valuation to $183 billion, as reported by PYMNTS. Founded in 2021 by former OpenAI researchers, Anthropic’s Claude models have become central to enterprise adoption, offering multimodal reasoning and compliance tools for regulated industries.
In 2025, the company introduced Claude Sonnet 4.5 and its Claude Agent SDK, extending its platform into developer tools and automation part of what PYMNTS called a strategy to turn Claude into an ecosystem. With these releases, Anthropic is positioning its models not just as chat interfaces but as infrastructure for building AI-native applications. The SDK allows developers to embed Claude’s reasoning capabilities into existing enterprise systems, while Sonnet 4.5 enhances multimodal understanding and real-time task execution. At the same time, PYMNTS reported that Microsoft has been exploring deeper ties with Anthropic as it reevaluates its reliance on OpenAI, signaling how cloud alliances are shifting as compute demands grow.
Amazon has already committed up to $8 billion to Anthropic and counts the company among the largest users of its custom AI chips. The talks with Google would further cement Anthropic’s multi-cloud approach, ensuring redundancy and access to the most advanced silicon. For Google, securing Anthropic as a long-term client could strengthen its position against Amazon and Microsoft in the cloud AI supply chain.
Source: https://www.pymnts.com/