Google Cloud signs multi-billion-dollar AI infrastructure deal with Thinking Machines Lab

Thinking Machines Lab, founded by former OpenAI executive Mira Murati in February 2025, has inked a multi-billion-dollar agreement with Google Cloud to expand its use of AI infrastructure, including systems powered by Nvidia's latest GB300 chips. The deal, valued in the single-digit billions, grants access to Google's AI systems and infrastructure services for model training and deployment, with Google noting it can support the startup's reinforcement learning workloads. This marks the first time Thinking Machines has partnered with a cloud services provider, though the arrangement is not exclusive. Earlier this year, the lab partnered with Nvidia, which included an investment from the chipmaker. Google has been actively securing cloud deals with AI developers, such as Anthropic's agreement earlier this month for tensor processing unit capacity with Google and Broadcom. Competition remains fierce, with Anthropic also signing a deal with Amazon this week for up to 5 gigawatts of capacity. Thinking Machines launched its first product, Tinker, in October 2025—a tool automating custom frontier AI model creation—after raising a $2 billion seed round at a $12 billion valuation. The startup is among the first Google Cloud customers to access GB300-powered systems, offering a 2X improvement in training and serving speed. Myle Ott, a founding researcher at Thinking Machines, stated that Google Cloud enabled record speed with reliability.