SAN FRANCISCO; Artificial intelligence company Anthropic announced a massive expansion of its collaboration with Google on Thursday, involving the purchase of up to one million Tensor Processing Units (TPUs) — specialised AI chips designed by Google — in a deal valued at tens of billions of dollars.
According to Anthropic, this move will allow the company to bring more than a gigawatt of new computing power online by the end of 2026, helping it meet rapidly growing demand for its Claude AI models. “This expansion ensures we can continue scaling to serve our more than 300,000 business customers worldwide,” said Anthropic’s Chief Financial Officer, Krishna Rao.
Founded in 2021 by former OpenAI employees, Anthropic was created out of concerns over responsible AI development. The company, backed by Amazon, quickly rose to prominence in the generative AI industry following the launch of ChatGPT by OpenAI in 2022.
Google Cloud CEO Thomas Kurian said the deal reflects the efficiency and strong performance Anthropic has experienced with TPUs over the years. “We continue to drive innovations that increase the capacity and cost-effectiveness of our TPUs,” he added.
Anthropic recently introduced its latest AI model, Claude Sonnet 4.5, described as the world’s most advanced for computer programming and capable of interacting with computers like a human. The company also plans to open its first office in India next year, tapping into the country’s booming digital market of nearly 900 million internet users.
Anthropic’s current valuation stands at around $183 billion, while OpenAI has reportedly reached an estimated $500 billion, making it the most valuable startup globally.
