Nvidia is in talks to invest up to $10 billion as an anchor investor in Anthropic’s planned initial public offering, which could raise as much as $100 billion at a valuation of roughly $2 trillion, according to a Reuters report that would make the listing the largest in history.
The talks are preliminary and terms could change, but the shape of the deal is taking form. Anthropic has selected Nasdaq for a potential listing, according to Business Insider, and the offering is expected to be completed before the US midterm elections in November. Bankers told potential investors in August that the company could seek to raise more than $100 billion, a figure that would exceed any IPO on record and value the five-year-old startup above Elon Musk’s SpaceX.
The numbers behind the valuation
Anthropic raised $65 billion in May at a post-money valuation of $965 billion. Its annualized revenue run rate passed $65 billion by the end of July, up from roughly $9 billion at the end of 2025. The company has told shareholders that adjusted operating income is expected to remain positive for a second consecutive quarter, according to the Financial Times, a detail that matters because most frontier AI labs are still burning cash.
The customer mix has concentrated at the top. When Anthropic announced its Series G in February, more than 500 business customers were each spending over $1 million on an annualized basis. That number passed 1,000 within two months, and the company’s run-rate revenue had already surpassed $30 billion when it announced its Google and Broadcom compute expansion in April.
Why Nvidia wants in
Nvidia already has deep financial and commercial ties to Anthropic. In November 2025, Nvidia said it would invest up to $10 billion in the company as part of a broader partnership under which Anthropic committed to purchase $30 billion of Microsoft Azure computing capacity powered by Nvidia chips. Anthropic is also backed by Amazon and Alphabet’s Google, its other major compute suppliers.
An anchor commitment in the IPO would deepen that relationship and give the listing a prominent early backer. It would also help gauge investor appetite for the valuations and capital requirements of frontier AI companies, a live question after a year in which AI infrastructure spending has dominated equity markets. Nvidia did not respond to a request for comment, and Anthropic declined to comment.
The compute bill keeps growing
Anthropic’s revenue growth is inseparable from its infrastructure commitments. In April, the company said it would commit more than $100 billion over a decade to Amazon Web Services and use more than one million of Amazon’s Trainium2 chips. The same month, it signed a deal with Google and Broadcom for multiple gigawatts of next-generation TPU capacity starting in 2027, with Broadcom’s regulatory filing specifying roughly 3.5 gigawatts accessible through the chipmaker.
The company trains and runs Claude across three hardware platforms, AWS Trainium, Google TPUs and Nvidia GPUs, which lets it match workloads to whichever chip suits them. That diversification is now a negotiating asset: every major chip supplier has a reason to keep Anthropic growing, and the IPO would give each of them a liquid stake in the outcome. Anthropic also expanded its Google Cloud relationship in October 2025 with access to up to one million TPUs, a deal worth tens of billions of dollars.
What the market is pricing
A $2 trillion valuation would put Anthropic among the most valuable companies in the world on the strength of roughly two years of hypergrowth. The math that justifies it is the run rate: if revenue sustained its trajectory from $9 billion to $65 billion, the multiple compresses quickly. The risk is that the trajectory does not hold, or that the compute costs required to sustain it outrun the revenue.
Anthropic’s consumption commitments are already the largest in the industry. The Broadcom filing noted that Anthropic’s consumption of the expanded capacity depends on its continued commercial success, an unusually candid caveat in a securities document. An IPO would transfer some of that execution risk to public shareholders.
Context: the sector’s biggest quarter
The listing would land in a market already reshaped by AI capital flows. Nvidia remains the largest AI processor maker, and its investments across the ecosystem, roughly 90 deals in 2025 and 60 more agreed by September 2026, function partly as demand generation for its own chips. Competitors have taken note: Google began recognizing revenue from direct TPU sales in the quarter ended June, and Amazon has opened talks to sell its Trainium chips to other data center operators. Microsoft, meanwhile, is preparing to unveil its Maia 300 chip with more than 300,000 units targeted for 2027 delivery.
Anthropic’s own disclosure that Claude now leads 26% of its internal AI research and development, published Thursday, adds another dimension to the valuation case. The company is telling investors its research pipeline is already a quarter automated, which is either an efficiency story or a risk story depending on how the market reads it.
The IPO also arrives amid an unresolved regulatory backdrop. The Senate failed to advance the CLARITY Act for crypto this week, and AI legislation has fared no better: the Sanders-Casar superintelligence ban proposal is a long shot, and no comprehensive AI package is expected this year. That leaves frontier labs operating in the same environment they have occupied all year, growing fast with federal rules largely unwritten.
Nothing is signed. The size, timing and anchor structure could all change before a filing. But the direction is set: the largest AI companies are moving from private capital markets to public ones, and Anthropic intends to set the price of entry.