Mira Murati's Thinking Machines Lab secures multi-billion-dollar Google Cloud deal for AI computing power

Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, signed a multi-billion-dollar agreement with Google Cloud to access Nvidia's latest GB300 chips—the most powerful AI processors available—for training and running large language models (AI systems that understand and generate text).

The GB300 chips deliver significantly higher performance than prior generations, meaning the lab can train larger AI models faster and run them more efficiently; this translates to faster responses in AI applications and the ability to build more capable systems with the same computational budget.

For businesses and researchers building AI products, this deal signals Google Cloud is competing aggressively to host cutting-edge AI workloads by securing exclusive access to the latest hardware. If you're evaluating cloud providers for AI projects, Google Cloud now has a credible claim as the platform where top-tier AI labs choose to build—which affects pricing pressure and feature roadmaps across the industry.

Thinking Machines Lab plans to use the infrastructure to develop AI agents (AI systems that can make decisions and take actions on their own) and other advanced models; the deal deepens Google's relationship with one of the most visible AI startups founded by an OpenAI alumnus.