Axle negotiates to lead a $1 billion financing round for Thinking Machines at a $40 billion valuation

Venture capital firm Axle is in advanced talks to lead a new $1 billion financing round for the AI lab Thinking Machines, in a deal that values the startup at no less than $40 billion, according to informed sources and corroborating reports. The negotiations come months after the company's leaders pursued a valuation near $50 billion late last year, with the upcoming round expected to settle below that ceiling while still representing one of the sector's largest valuations relative to the company's age.

Thinking Machines was founded early last year by Mira Murati, the former chief technology officer of OpenAI, and quickly built financial flows that exceeded an annual revenue rate of $100 million. Nevertheless, a valuation of roughly $40 billion reflects an exceptionally high revenue multiple even by Silicon Valley market standards, highlighting investors' bet on combining research efficiency with the development of specialized inference platforms.

The company's business model moves away from selling individual subscriptions and focuses on computing fees and the adaptation of open models.In July, the company launched the open-weight “Inclining” model and linked it to the “Tinker” platform, which generates revenue by charging computing fees based on the actual usage volume for model calibration and training on enterprises' proprietary data, allowing companies to own their adjusted weights without relinquishing the privacy of their internal data.

The company raised $2 billion in its first seed round at a $12 billion valuation, in a round led by Andreessen Horowitz and participated in by Nvidia, Google Ventures, Light Speed and Conviction Partners. Despite the investment confidence built on Murati's professional background and the founding research team, the lab recently saw the departure of prominent figures and the return of some co-founders to OpenAI, including Lillian Weng and Luke Metz, putting the new round under scrutiny regarding the stability of the leadership team and its ability to endure.

This shift in financing and pricing structure requires precise recalculations for infrastructure and data teams in the Gulf and the Arab region.Thinking Machines' reliance on computing fees to adapt open weights such as Inclining on dedicated platforms offers a practical alternative for entities that fear dependence on closed APIs from major U.S. firms, especially in banking and government sectors that require data to remain local. If you manage your institution's computing and AI budget, following these models gives you clarity on the cost of building specialized models on your data versus direct cloud-call expenses, and precisely indicates when investing in adaptation platforms becomes an economically viable option.