Eight private AI companies carry both a valuation of at least $10 billion and a reported revenue run rate of at least $100 million, by our count from FourWeekMBA’s AI landscape dataset and our published coverage. Together they are valued at about $1.6 trillion, by our arithmetic, and OpenAI is about 78% of that.
The median company in the list is valued at $31 billion, and the median valuation is about 32x the run rate, by our arithmetic. The list was prompted by a chart Deedy Das posted on X on the same theme; the list and every figure here are ours, each linked to the piece that reported it.
Business Pill · Run-Rate Revenue
A one-minute explainer of run-rate revenue. It teaches the general idea only and does not discuss this list.
The key insight: As we read the list, the value sits with one company: OpenAI is about 78% of the combined $1.6 trillion, yet at about 25x its multiple is below the median of about 32x, by our arithmetic.
OpenAI: a $1.25 trillion valuation on a reported $15 billion raise (CNBC, citing sources, 12 November); an annual run rate of roughly $50 billion (CFO Sarah Friar to Yahoo Finance, reported 10 November). Valuation to run rate: about 25x, by our arithmetic. Our coverage, run rate
Databricks: valued at $190 billion; a $7 billion revenue run rate, both as written by Horowitz in October. Valuation to run rate: about 27.1x, by our arithmetic. Our coverage
Cognition: a $48 billion valuation at its September Series E (Fortune); a crossing of $1 billion in annualized revenue run rate (Robinhood release). Valuation to run rate: about 48x, by our arithmetic. Our coverage
Thinking Machines Lab: raising at a $40 billion pre-money valuation (Business Insider, 6 November); a run rate of over $100 million in September (one source cited by TechCrunch, not confirmed). Valuation to run rate: up to about 400x, by our arithmetic. Our coverage
ElevenLabs: $22 billion in its employee tender offer (company blog, 30 September); global revenue that has crossed a $600 million annual run rate (CNBC-TV18). Valuation to run rate: about 36.7x, by our arithmetic. Our coverage
Mercor: a $20 billion valuation at its Series D (company release, 10 November); $3 billion in run-rate revenue, a claim by its CEO that is not independently verified. Valuation to run rate: about 6.7x, by our arithmetic. Our coverage
fal: a $16 billion valuation in a $1.3 billion round led by Bessemer; $800 million in annualized revenue this year (a source to Bloomberg). Valuation to run rate: about 20x, by our arithmetic. Our coverage
Harvey: a $15.5 billion valuation announced on 9 September; a crossing of US$400 million in annual recurring revenue (Ontario Teachers’). Valuation to run rate: about 38.8x, by our arithmetic. Our coverage
Valuation as a multiple of reported revenue run rate for the private AI companies in our list; figures as reported, multiples our arithmetic. Thinking Machines Lab is not shown.
How We Built the List
A company qualifies if it is private, builds AI products or infrastructure, and has a latest reported valuation of $10 billion or more and a latest reported run rate, annualized revenue or annual recurring revenue of $100 million or more. Each valuation is the one its source gives, pre- or post-money; rounds still in talks are marked as reported.
Listed companies are left out. Companies with only one of the two figures in our coverage are listed separately below, as are figures that are third-party estimates.
Why Anthropic Is Not Counted
Anthropic has filed to go public: its IPO site says the share price is expected to be set on 18 November and trading to begin on 19 November, as we reported. We therefore leave it out of the count.
On its last private mark it would qualify: a $965 billion post-money valuation in May and run-rate revenue of about $81 billion on 31 October, per its amended prospectus. That is about 11.9x, by our arithmetic, and adding it would take the combined value to about $2.57 trillion.
Companies That Meet One Bar
Valuation without a run rate in our coverage: Shield AI at a $28.8 billion post-money valuation, and Fireworks AI at $17.5 billion in July. Applied Intuition is in talks at about $30 billion, Bloomberg reported, and booked $474 million of revenue the year before, an annual figure rather than a run rate.
Run rate without a $10 billion valuation: micro1 says it passed a $1 billion run rate and priced its Series B at $4 billion, and Legora reports more than $200 million of annual recurring revenue at an $8.8 billion valuation.
Estimate only: wire headlines put DeepSeek’s pre-money valuation at $74 billion, but its $1.0 billion revenue figure is an Epoch AI estimate, not a company or outlet-sourced number. Outside the AI definition: Ramp, which Bloomberg calls a fintech startup, has a reported valuation of about $60 billion and says annualized revenue has passed $2 billion.
What the Multiples Show
The lowest multiple in the list is Mercor’s, at about 6.7x, and it rests on its CEO’s unverified run-rate claim. The highest is Thinking Machines Lab’s: up to about 400x, on a single-source figure from September that gives only a floor.
Excluding those two, the other six sit between about 20x for fal and 48x for Cognition, by our arithmetic. OpenAI, at about 25x, is below the median even though it is about 78% of the combined value.
The Structural Read
As we read the spread, the multiples fall into a band of about 20x to 48x for six of the eight; the two outliers rest on the least certain run-rate figures in the list, a CEO claim and a single unconfirmed source.
As we read the one-bar names, the list is likely to change as figures are reported: several companies at $10 billion or more have no run rate in our coverage, and micro1 and Legora report run rates above $100 million at valuations below $10 billion.
Ontario Teachers’ Pension Plan on Harvey, 24 September 2026, as quoted in our coverage
“recently crossed US$400M in annual recurring revenue after only four years”
Three Implications
FOR READERS COMPARING MARKS Run-rate definitions differ by company (ARR, annualized revenue, run-rate revenue), so each figure is shown with its source.
FOR ANTHROPIC On its May post-money valuation and its run-rate revenue of about $81 billion, it would sit at about 11.9x, by our arithmetic.
FOR THE WEEKLY REFRESH New valuations or run rates reported for one-bar companies such as Shield AI, Fireworks AI or micro1 would change the count.
The framework says: “Every headline, every valuation move, every strategic decision in the cycle lives inside one of these nine boxes.”
As we read it, the eight companies sit in different layers: models (OpenAI, Thinking Machines Lab), data platforms (Databricks), inference (fal), training data (Mercor) and applications (Cognition, Harvey, ElevenLabs). The framework does not discuss this list.
What Is Not Established
Run-rate definitions differ: some companies give annual recurring revenue, others annualized revenue or run-rate revenue, and the dates run from September to November. Several figures come from a single source or from rounds still in talks, and the list is limited to what our coverage has sourced.
It is not a complete census: Bessemer’s AI 100 ranks 100 private AI companies with a total value of $3.6 trillion, and companies without both figures in our reporting are not counted. We will refresh the list weekly as new figures are reported.
This Week In The AI Supercycle
AI Lost Its Central Banker, the Lab Oligopoly, and the Future of Open Source
This week’s issue: the Nvidia leadership change, read as the AI build-out losing its central banker: who now finances the GPU landlords, backstops the data centres and sponsors open models, and why two labs are becoming the bank.
By our count, eight private AI companies clear both a $10 billion valuation and a $100 million run rate. Together they are worth about $1.6 trillion, at a median of about 32x run rate, by our arithmetic. Anthropic would also qualify on its last private marks; its IPO site expects pricing on 18 November.
A note on sourcing. Every company figure comes from a FourWeekMBA piece linked in the list, as reported there by the company or the named outlet, and is held in our AI landscape dataset; the list method, totals, medians and multiples are ours. Deedy Das’s chart on X prompted the list; none of its figures are used.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
