Thinking Machines Lab is reportedly raising $5 billion in a round led by Accel at a pre-money valuation of $40 billion, according to Business Insider1,2. The deal could close as soon as this week.
The valuation more than triples the $12 billion at which Thinking Machines Lab raised a $2 billion seed in July 2025, though it sits roughly 20% below the $50 billion figure discussed in late 2025. If the round closes, the company will have raised about $7 billion in disclosed funding, excluding an undisclosed March investment from Nvidia Corp..
Revenue and valuation gap
Revenue remains modest relative to the price tag. A source close to the numbers told TechCrunch in September that annual revenue exceeded $100 million, which would put the valuation at up to 400 times revenue. For comparison, OpenAI's $1.25 trillion valuation is around 25 times its September run rate.
The company generates revenue through Tinker, a product that lets organizations train their own AI models on proprietary data; Thinking Machines Lab handles the computing and collects usage fees.
Investor roster and co-founder departures
Accel participated in the seed round alongside Andreessen Horowitz, Nvidia, AMD, ServiceNow, Cisco, and Jane Street. Thinking Machines Lab was founded in San Francisco in February 2025 by Mira Murati, former OpenAI chief technology officer. Between October 2025 and January 2026, three of Murati's co-founders left for Meta and OpenAI.
In March, Thinking Machines Lab and Nvidia announced a multi-year partnership to deploy at least one gigawatt of Nvidia's Vera Rubin systems, with first deployment scheduled for early 2027.
Model releases
The company's first open-weight model, Inkling, launched on July 15. Inkling 2 Small came out on November 5, and Thinking Machines Lab intends to release the full weights under an Apache 2.0 license the following week, pending external safety testing. Inkling 2 Small beats Mistral Large 4 on the Artificial Analysis index, according to the company.
ANALYSIS The 400-times-revenue multiple, even by frontier-AI standards, stretches well beyond the ratios carried by better-established labs, placing the bet squarely on Thinking Machines Lab's compute partnerships and model roadmap rather than current commercial traction.