Anthropic is in talks to bring Nvidia in as an anchor investor for what could be the largest IPO in history, targeting a raise of up to $100 billion at a valuation near $2 trillion, Reuters reported on September 11. Two people familiar with the matter said Nvidia is considering investing up to $10 billion. The plans remain under discussion and could change, the sources cautioned, requesting anonymity because the discussions are confidential.
The listing would top every technology debut Wall Street has seen. Anthropic filed confidentially in June and has been building toward a roadshow, with the IPO expected to complete before the US midterm elections in November. Its last private round, a $65 billion raise in May, valued the company at $965 billion, and its annualized revenue run rate had climbed above $65 billion by the end of July, up from about $9 billion at the end of 2025, according to the company.
Why Nvidia would write the check
The relationship between the two companies is already deep. In November 2025, Nvidia said it would invest up to $10 billion in Anthropic as part of a broader partnership under which Anthropic committed to buy $30 billion of Microsoft Azure computing capacity powered by Nvidia chips. Anthropic relies heavily on Nvidia GPUs while trying to diversify its suppliers, because demand for Claude has strained its available computing capacity.
An anchor investment in the IPO would deepen that tie at exactly the moment Anthropic is renegotiating its hardware future. In April, the company said it would commit more than $100 billion over a decade to Amazon’s AWS, using more than 1 million of Amazon’s Trainium2 chips. It has also agreed with Google and Broadcom to add multiple gigawatts of TPU capacity. Anthropic is building an in-house team to design custom chips tailored to Claude, part of an effort to gain control over hardware costs as its revenue surges.
For Nvidia, the calculus is defensive as much as offensive. Its largest customers are becoming its most credible alternatives, and taking a stake in the biggest IPO of the cycle keeps Nvidia inside the cap table of the company it sells to. The move would echo a pattern from the last two years: Nvidia investing in the labs that buy its chips, and those labs using the investment to buy more chips. The company has already put more than $50 billion into frontier labs, by its own account on a recent earnings call, alongside a $6 billion deal with coding startup Poolside to build open models and a $20 billion licensing agreement with AI chip startup Groq.
A valuation the market will have to digest
A $2 trillion valuation would put Anthropic ahead of nearly every public company in the world on day one. The revenue supports part of the story. A run rate above $65 billion, growing from $9 billion in seven months, is growth almost no public company has matched. But the price is also a bet that frontier model revenue keeps compounding while compute costs, competition from OpenAI and Google, and regulatory pressure all cut the other way.
The IPO would also mark a shift in how AI labs finance themselves. OpenAI has raised through SoftBank-led rounds and debt, including a $10 billion loan arranged by SoftBank in August. Anthropic chose the public markets, which puts its books under scrutiny and its valuation in the hands of daily traders rather than a handful of sovereign funds. An anchor investor like Nvidia would de-risk the offering by signaling conviction at the top of the book, and it would give retail investors who missed the private rounds a way in.
Wall Street braces for the largest IPO ever
Bankers have spent the summer preparing for a wave of AI listings, and Anthropic’s filing is the centerpiece. The scale matters beyond the headline. A $100 billion raise would give Anthropic a war chest that changes the compute market by itself, letting it pre-purchase chips and data center capacity the way oil majors lock up drilling rights. Suppliers from Amazon to Google to Nvidia have all negotiated decade-long commitments to keep Anthropic’s models running, and the IPO proceeds would give the company leverage in every future negotiation.
Investors will watch three numbers when the prospectus lands: gross margin after compute costs, the split between consumer and enterprise revenue, and how much of the $65 billion run rate comes from multi-year commitments versus monthly subscriptions. Each has moved the valuation conversation for every AI listing so far this cycle, and Anthropic’s disclosure choices will set the template for the labs that follow it to market.
The buyer base matters too. Anthropic’s private rounds pulled in an unusual mix of sovereign wealth funds from the Gulf, hedge funds like Coatue and D.E. Shaw, and strategics including Amazon, Google, Microsoft and Nvidia itself. Converting that shareholder register into public float will test how much of the demand was conviction and how much was access. The underwriters will want anchor commitments locked before pricing, which is precisely the role the Nvidia talks address.
The timing also collides with a live policy fight. Lawmakers in Washington are weighing AI regulation calls that grew louder after a researcher’s extinction warning, and a dozen OpenAI and Anthropic insiders have asked the labs to slow down. A company debuting at $2 trillion while its own employees debate pacing will face questions no roadshow script fully answers.
Neither company has commented publicly. If the talks hold, the roadshow opens in mid-October, and the largest IPO in history will price a few weeks before American voters go to the polls.