Nvidia is negotiating a 2.500 millones de dólares investment in artificial intelligence startup Thinking Machines Lab, founded by former OpenAI executive Mira Murati. The discussions coincide with a separate 1.000 millones de dólares funding round that could value the startup at approximately 40.000 millones de dólares, building on an existing partnership between the two companies.
Nvidia and Mira Murati’s Thinking Machines Lab Expand Partnership
Nvidia is in talks to invest roughly 2.500 millones de dólares in Thinking Machines Lab, according to reporting from Investing.com citing The Information. The potential capital injection builds upon a pre-existing relationship between the chipmaker and the artificial intelligence startup founded by former OpenAI chief technology officer Mira Murati. Earlier in March, the two companies announced a multiyear strategic partnership centered on deploying at least 1 gigavatio of next-generation Nvidia Vera Rubin systems, with deployment targeted for early next year.
That foundational agreement also encompassed an effort to design training and serving systems specifically for Nvidia architectures while broadening enterprise access to frontier models. Jensen Huang, founder and chief executive officer of Nvidia, praised the startup’s team during the partnership announcement.
Huang added that Thinking Machines has assembled a world-class team to push the boundaries of artificial intelligence. Murati echoed the sentiment regarding the hardware foundation underpinning her firm’s research.
Murati noted that the alliance accelerates the startup’s capacity to build customizable artificial intelligence systems that people can shape and make their own.
Venture Capital Backing and Valuation Adjustments at Thinking Machines
Alongside the prospective Nvidia investment, Thinking Machines is concurrently negotiating to secure at least 1.000 millones de dólares in a new financing round led by venture capital firm Accel. This round values the artificial intelligence laboratory at approximately 40.000 millones de dólares, according to market sources.
The current valuation target reflects a downward adjustment from late last year, when the startup explored raising capital at a valuation exceeding 50.000 millones de dólares. Despite the modest pullback in valuation targets, the company reports an annual run rate of revenue exceeding 100 millones de dólares. That figure represents a remarkably high revenue multiple relative to its corporate valuation.
The young company, which has been operational for less than two years, previously completed a 2.000 millones de dólares financing round that valued the business at 12.000 millones de dólares. That early-stage raise was led by Andreessen Horowitz with participation from Nvidia, GV, Lightspeed, and Conviction Partners, drawing heavy backing due to the industry reputation of Murati and the researchers who departed OpenAI to join her. Since that formation, the enterprise has experienced high-profile departures, including cofounders Lilian Weng and Luke Metz returning to OpenAI.
Inkling Launch and Commercial Strategy
Commercial operations at Thinking Machines gained traction following the July release of Inkling, an open-weight model designed to generate revenue through computing fees based on usage. The platform, known as Tinker, enables clients to adapt advanced models directly to proprietary data.
The company’s commercial model relies on usage-based computing tariffs. By pairing open-weight foundation models with specialized fine-tuning infrastructure, Thinking Machines positions itself across both frontier model research and enterprise deployment layers.
Nvidia’s Broader Ecosystem Strategy and Market Scrutiny
The ongoing talks with Thinking Machines highlight a broader corporate strategy by the chipmaker to deploy capital across the entire artificial intelligence ecosystem. According to data from PitchBook cited by Forbesargentina, Nvidia has invested approximately US$ 53.000 millones across 170 separate operations. The hardware giant participated in nearly 67 venture capital funding rounds in 2025 alone, a sharp increase from 54 rounds in 2024 and just 12 in 2022.
Nvidia’s venture capital arm, NVentures, closed 30 individual transactions during 2025. The company’s portfolio spans foundational model developers including OpenAI, Anthropic, Mistral, xAI, and Cohere, alongside infrastructure providers like CoreWeave, Nscale, and Nebius, as well as ventures in autonomous driving, robotics, quantum computing, and nuclear fusion.
This financing strategy has drawn comparisons from short sellers such as Jim Chanos and Michael Burry to the financing arrangements that preceded the collapse of telecommunications equipment maker Lucent during the dot-com bubble. Critics suggest that vendor financing could entangle hardware sales with investment activity.
In response, Nvidia issued a seven-page memorandum to Wall Street analysts defending its practices. The company noted that its clients typically settle invoices within 53 days rather than years, and that its commercial revenues remain independent of its venture investments. Market analysts point out a structural difference: unlike historical precedents where equipment vendors financed customers lacking alternative capital, Nvidia’s primary customers—hyperscale cloud providers including Microsoft, Google, Meta, and Amazon—are deploying hundreds of billions of dollars in independent capital toward artificial intelligence infrastructure.
