Anthropic, a prominent player in the artificial intelligence sector, revealed on Thursday its plans to significantly enhance its collaboration with Google through an agreement valued in the tens of billions of dollars. This partnership will see Anthropic acquire up to one million Tensor Processing Units (TPUs), which are specialized integrated circuits developed by Google, according to the San Francisco-based startup.
This strategic expansion aims to bring over a gigawatt of additional computing power online by the close of 2026. Such an increase is crucial for Anthropic as it seeks to address the “exponentially growing demand” for its Claude AI models, as noted by Chief Financial Officer Krishna Rao in a statement. Currently, Anthropic serves more than 300,000 business clients.
Founded in early 2021 by former employees of OpenAI, Anthropic emerged from concerns that their previous employer, led by CEO Sam Altman, was not taking sufficient measures to mitigate the potential risks associated with its AI models. With backing from Amazon, the startup quickly positioned itself among the leaders in the generative AI field, especially following the release of OpenAI”s ChatGPT in November 2022, which sparked a competitive race in AI advancements.
Although Anthropic has not yet matched OpenAI”s user base and brand recognition, it has been recognized as a top contender in generative AI, particularly in the realm of computer programming. Meanwhile, Google is investing heavily in AI technologies to compete with other industry giants, including OpenAI, Microsoft, and Amazon, in the ongoing battle for leadership in artificial intelligence.
The TPUs that Anthropic plans to utilize offer a competitive alternative to the highly sought-after GPUs from Nvidia. “Anthropic”s decision to significantly increase its use of TPUs highlights the strong price performance and efficiency our teams have experienced with these units over the years,” stated Thomas Kurian, head of Google Cloud. “We continue to innovate and enhance the efficiency and capacity of our TPUs.”
Recently, Anthropic introduced its latest generative AI model, Claude Sonnet 4.5, claiming it to be the most advanced for computer programming tasks. This model is also acknowledged for its sophistication in allowing AI assistants to operate computers in a manner similar to humans.
Furthermore, Anthropic announced its intention to establish an office in India next year, as global players in generative AI seek greater influence in the world”s most populous nation. The demand for AI solutions has surged in India, which is expected to surpass 900 million internet users by the end of the year, driven by increased adoption from both businesses and individuals.
Currently, Anthropic is valued at approximately $183 billion, in contrast to OpenAI, whose valuation reportedly surged to $500 billion following a recent private share sale, potentially making it the most valuable startup globally.
With a diverse workforce of 2,400 employees from 100 different nationalities, AFP provides fast, comprehensive, and verified reporting on the significant issues that impact our daily lives.
