Lambda, the AI cloud company, told investors this spring that its backlog of committed orders stood at $15 billion. By September the number had more than tripled, to $50 billion. That jump is the backdrop to the company’s latest move: raising up to $4 billion in what would be its final private round before an initial public offering.
The Wall Street Journal reported the fundraising on Tuesday, citing people familiar with the matter. Blackstone and Coatue Management are leading the round, which would value Lambda at $14.5 billion before the new money arrives. Reuters followed. The company has told investors a 2027 public debut is the target, though the timing depends on execution and the state of the markets.
Much of the growth in that backlog rests on a single customer. In late August Lambda signed a roughly $35 billion cloud-services agreement with Anthropic, the maker of the Claude AI models, under which Anthropic will lease Nvidia hardware housed in a Texas facility operated by Hut8. Strip out that one contract and the three-month jump in demand looks less like a broad surge than one enormous commitment.
That concentration cuts both ways. It is why Lambda can raise at a valuation that has climbed sharply from a year ago, and it is why investors are asking how much of the company’s future rests on one customer’s ability to keep paying. Anthropic is itself racing to fund its expansion, including a reported effort to secure tens of billions of dollars in chip financing. The AI boom rewards companies that grow fast and punishes those whose growth slows.
Lambda’s rise tracks the wider neocloud boom. The company rents out access to Nvidia chips and server architecture, and it uses Nvidia hardware exclusively — a loyalty that has made Nvidia an investor. Nvidia took part in Lambda’s $480 million Series D round in February 2025, and the chipmaker’s backing signals to the market that Lambda sits close to the source of the hardware every AI company is scrambling to get.
The valuation tells its own story. As recently as November 2025, a round led by TWG Capital valued Lambda at $5.9 billion after the new investment. A $14.5 billion pre-money valuation nine months later is a steep climb, and it reflects the arithmetic of the current AI build-out: customers are committing to enormous multi-year contracts, and the companies that win those contracts are being marked up accordingly.
Underpinning all of it is a build-out with no close historical parallel. Data-center capital spending across the industry is projected to reach roughly $3 trillion a year by the end of the decade, by some estimates, and the neoclouds are the companies converting that money into racks of servers and megawatts of power. Every new contract they sign is a promise to spend heavily before the revenue arrives.
The financing model behind these firms is unusual, too. Because their revenue is contracted years in advance, lenders and investors have been willing to extend credit against future cash flows in a way few young companies enjoy. Lambda’s peers have used that willingness to raise debt secured against their data centers, and the equity rounds are only one leg of a capital structure built on the assumption that AI demand keeps compounding.
Lambda is not alone in that trade. CoreWeave and Nebius, two other Nvidia-backed cloud providers, have turned to public markets to fund their data-center expansions, and all three now depend partly on the health of their stock prices to keep building. Lambda had been expected to go public this year but pushed the debut back amid choppy markets. A pre-IPO raise of $4 billion is a way to keep building without waiting for the public window to open.
For Nvidia, the neoclouds are a channel as much as a customer base. Every dollar they raise flows toward the purchase of Nvidia’s accelerators, which is why the chipmaker has spread investments across Lambda, CoreWeave, Nebius and others. The strategy hedges its own exposure while steering capital toward the hardware its ecosystem depends on.
The raise also sets a tone for the IPO itself. A private round at $14.5 billion gives public-market investors a reference point, and a valuation that has risen so fast invites scrutiny of how much of it is durable. Analysts who follow the sector said the question is the same one hanging over the entire AI cloud business: whether the customer commitments that justify these valuations will turn into cash over the years those contracts run.
For Lambda, the practical answer is to keep raising while it can. The company has told investors it needs the capital to expand, and the market, for now, is willing to supply it. The backlog number — $50 billion — is the headline, but the real test will come when those commitments are measured against the cost of the chips, power and buildings required to honor them.
