Artificial intelligence startup Thinking Machines Lab, founded by OpenAI co-founder Mira Murati, has signed a multi-year strategic partnership with semiconductor leader Nvidia to expand its computing infrastructure for advanced AI research.
The agreement includes the deployment of at least one gigawatt of Nvidia’s Vera Rubin AI systems, beginning in 2027. As part of the partnership, Nvidia is also making a strategic investment in the company. Financial terms of the deal were not disclosed.
Founded in February 2025, Thinking Machines Lab has raised more than $2 billion from investors including Andreessen Horowitz, Accel, Nvidia, and the venture arm of AMD, valuing the company at over $12 billion. The startup is focused on developing AI models designed to produce reproducible results and released its first product, the Tinker API, in October.
According to the companies, the collaboration will also support the development of training and inference systems optimized for Nvidia architecture, reflecting growing demand for large-scale AI computing infrastructure.
