AI company Anthropic is reportedly in conversation with Google regarding a major cloud computing contract.

The deal is said to be valued in the high tens of billions of dollars, according to a report from Bloomberg, which cites people familiar with the matter.

The contract would see Anthropic significantly expand its use of Google Cloud and its in-house tensor processing units (TPUs) AI chips.

Both companies declined to comment to Bloomberg regarding the deal.

Anthropic was founded in 2021 by former OpenAI employees and directly competes with OpenAI through its Claude large language models. Google is a notable investor in the company, having invested around $2bn in 2023 and another $1bn in 2025.

In March of this year, court documents revealed Google held a 14 percent stake in the AI company.

Despite this, Anthropic has long been closely associated with rival Amazon Web Services (AWS).

Amazon has invested some $8 billion into Anthropic, which made it the company's "primary cloud and training partner," and is using "AWS Trainium and Inferentia chips to train and deploy its future foundation models" alongside Nvidia GPUs.

AWS is developing a supercluster of its Trainium2 UltraServers dubbed "Project Rainier" for Anthropic, which will feature "hundreds of thousands" of the chips interconnected with third-generation, low-latency petabit-scale EFA networking, and spread across multiple data centers in the US.

Anthropic uses both Amazon and Google's cloud services, announcing that it would deploy TPU v5e chips to train generative AI models back in 2023.