Accelerating Thinking Machines Lab’s frontier AI development

San Francisco / Sydney – Today we've announced a multi-year agreement with Thinking Machines Lab, an AI research and product company, which will become an anchor customer for the Firmus AI-Cloud Region in Asia.

Under the agreement, Thinking Machines Lab will train and deploy its next generation of AI models on Firmus AI Factory infrastructure powered by NVIDIA GB300 systems.

Founded in 2025, Thinking Machines Lab is a research lab and product company whose mission is to build AI that extends human will and judgment. This calls for AIs as diverse and distributed as people themselves are. The company trains strong models and builds tools that let people fine-tune models with their unique knowledge and keep adapting them as that knowledge evolves. It develops interaction models that handle live, multimodal interaction natively, so people can interrupt and correct and change their minds aloud. It publishes research and recipes that put these capabilities within reach of more people.

The Firmus AI-Cloud Region is built on the company’s AI Factory architecture, integrating networking, compute and software into a unified platform designed to support frontier AI workloads. Firmus delivers the performance and resilience frontier AI needs through the Firmus HyperCube, integrating liquid cooling, power and mechanical systems to reduce cooling and energy demand.

“We’re delighted to welcome Thinking Machines Lab as an anchor customer for our AI-Cloud Region in Asia,” said Trent Viengsone, Vice President, North America, Firmus Technologies. “Thinking Machines Lab’s work sits at the leading edge of AI research and development, and this agreement reflects the growing demand for purpose-built AI infrastructure capable of supporting the next generation of models. We look forward to helping Thinking Machines Lab scale its ambitions on Firmus AI Factory infrastructure.”

“Our Model-to-Grid approach recognises that networking, cooling and power are as important as compute for efficiency,” said Dr Daniel Kearney, Chief Technology Officer at Firmus Technologies. “By designing these systems together, we give Thinking Machines Lab additional capacity to train, deploy and scale advanced AI systems.”