Thinking Machines is back at the fundraising table, and the numbers tell a story that’s either audacious or unhinged depending on who you ask. Mira Murati’s AI lab is in talks to raise $1 billion at a valuation of at least $40 billion, with existing backer Accel reportedly leading the round. That’s a step down from the $50 billion the company was reportedly chasing late last year — but the real eyebrow-raiser is the revenue math. Sources say Thinking Machines is running at just over $100 million in annual revenue. That puts the valuation-to-revenue multiple somewhere around 400x, a figure that makes even the frothiest SaaS valuations of 2021 look restrained.
The company’s pitch centers on Inkling, an open-weight model introduced in July that makes money through usage-based compute fees on its Tinker platform, where customers adapt models on proprietary data. It’s a clever wedge — open weights to lure developers, metered compute to capture revenue — but $100 million in run rate doesn’t exactly scream category dominance. Not yet, anyway.
This is the second act for a startup that debuted with one of the largest seed rounds in history: $2 billion at a $12 billion valuation, led by Andreessen Horowitz with Nvidia, GV, Lightspeed, and Conviction Partners piling in. That bet was placed almost entirely on Murati’s track record and the OpenAI researchers who followed her out the door. Since then, the narrative has gotten bumpier. Several high-profile co-founders, including Lilian Weng and Luke Metz, have already boomeranged back to OpenAI. Investors who bought the talent story are now being asked to price the product story.
Accel leading the round is notable — the firm has a reputation for discipline that feels almost contrarian in today’s AI funding environment. If they’re willing to anchor at $40 billion despite the revenue gap, it signals conviction that Thinking Machines can scale its compute-fee model fast. But at what multiple do you draw the line? A $40 billion price on $100 million in revenue assumes a growth trajectory that makes even OpenAI’s numbers look modest. Murati has the pedigree and the platform. What she doesn’t have yet is proof that developers will pay enough for Tinker to justify the sticker shock.
