AI company Anthropic has entered a multi-billion dollar agreement with Google, gaining access to a million TPUs. The deal expands Anthropic's computational resources for AI development.
On October 24, 2025, Anthropic announced a significant partnership with Google, securing a multi-billion dollar deal that provides access to one million Tensor Processing Units (TPUs). This agreement builds on previous collaborations, allowing Anthropic to scale its AI infrastructure substantially.
TPUs are Google's specialized hardware for machine learning tasks, and this access is described as enabling Anthropic to handle even more advanced AI workloads. The deal underscores the growing reliance on cloud computing power in the AI sector, where high-performance resources like TPUs are essential for training large models.
While specific financial details and timelines beyond the announcement were not disclosed in the report, the partnership highlights the competitive landscape of AI hardware provision. Anthropic, known for its Claude AI models, continues to prioritize safe and reliable AI development through such strategic alliances.