Thinking Machines Lab secured a US$2bn seed round before releasing a single product. Eighteen months later, the San Francisco lab operates with two products, a revenue run-rate reportedly exceeding US$100m, and at least a gigawatt of NVIDIA compute on order.
Mira Murati Leads the San Francisco AI Lab
Former OpenAI CTO Mira Murati founded the lab on the premise that companies want AI they can shape themselves rather than rent as a finished product. That strategy centers on two main offerings: Tinker, a fine-tuning API, and Inkling, a 975-billion-parameter open-weights model. Investors are currently weighing a valuation of US$40bn for the startup as it scales its operations.
The lab’s infrastructure expansion relies heavily on hardware procurement. Speaking about the partnership with NVIDIA to deploy at least one gigawatt of Vera Rubin systems, CEO Mira Murati stated: This partnership accelerates our capacity to build AI that people can shape and make their own, as it shapes human potential in turn.
NVIDIA systems and Inkling model drive US$40bn valuation
The deployment of at least a gigawatt of NVIDIA Vera Rubin systems positions the startup among major enterprise AI infrastructure buyers.
Plans for Tinker and the 975-billion-parameter Inkling model led investors to consider a US$40bn valuation for Thinking Machines Lab.
What products and valuations define Thinking Machines Lab?
What products does Thinking Machines Lab currently offer?
The lab offers Tinker, a fine-tuning API, and Inkling, a 975-billion-parameter open-weights model designed for customizable enterprise deployment.
How much compute has the lab ordered from NVIDIA?
The San Francisco startup has placed an order for at least one gigawatt of NVIDIA Vera Rubin systems to power its infrastructure expansion.
What is the projected valuation of the startup?
Investors are weighing a valuation of US$40bn for Thinking Machines Lab following its rapid revenue growth and product launches.
Worth a look