Exclusive: Google deepens Thinking Machines Lab ties with new multi-billion-dollar deal
In a significant move within the tech industry, Google has announced a new multibillion-dollar agreement with Thinking Machines Lab, a startup founded by former OpenAI executive Mira Murati. This partnership aims to enhance the use of Google Cloud’s AI infrastructure, leveraging advanced systems powered by Nvidia’s latest GPUs.
Details of the Agreement
The deal, valued in the single-digit billions, provides Thinking Machines Lab access to Google’s cutting-edge AI systems built on Nvidia’s new GB300 chips. This collaboration includes essential infrastructure services that will support model training and deployment, marking a pivotal step for both companies in the rapidly evolving AI landscape.
Google’s Strategy in AI Cloud Services
Google has been actively pursuing cloud agreements with AI developers, aiming to integrate its AI computing offerings with other cloud services such as:
- Storage solutions
- Kubernetes engine
- Spanner, its database product
Earlier this month, another AI company, Anthropic, signed a deal with Google and Broadcom for substantial tensor processing unit (TPU) capacity, showcasing the competitive nature of the AI cloud market.
Thinking Machines Lab’s Background
Founded in February 2025 by Mira Murati, Thinking Machines Lab quickly gained attention after raising a $2 billion seed round at a valuation of $12 billion. The startup has maintained a level of secrecy around its operations but launched its first product, Tinker, in October 2025. Tinker is designed to automate the creation of custom frontier AI models.
Insights into Thinking Machines Lab’s Developments
The recent agreement with Google sheds light on the technological advancements Thinking Machines is pursuing. Google emphasized its capability to support the startup’s reinforcement learning workloads, which are integral to Tinker’s architecture. Reinforcement learning has been a crucial training approach that has led to significant breakthroughs in AI, particularly at renowned labs like DeepMind and OpenAI.
Performance Enhancements with Google Cloud
Thinking Machines is among the first customers to utilize Google Cloud’s GB300-powered systems, which reportedly offer a two-fold improvement in training and serving speed compared to previous GPU generations. Myle Ott, a founding researcher at Thinking Machines, expressed enthusiasm about the partnership, stating, “Google Cloud got us running at record speed with the reliability we demand.”
The Competitive Landscape
The AI cloud services market is highly competitive, with major players like Amazon also vying for dominance. Just this week, Anthropic secured a new agreement with Amazon for up to 5 gigawatts of capacity for training and deploying its AI model, Claude. This competition underscores the urgency for companies like Thinking Machines to establish strong partnerships early in their development.
Conclusion
The collaboration between Google and Thinking Machines Lab represents a strategic alliance that could shape the future of AI development. As both companies continue to innovate and push the boundaries of technology, the implications of this partnership may extend beyond their immediate projects, influencing the broader landscape of AI and cloud computing.
Note: The information in this article is based on sources available as of April 2026 and is subject to change as new developments occur in the tech industry.
