Stay connected with KayaToday—follow us on Instagram and Facebook for the latest news and reviews delivered straight to you.
In February of this year, Anthropic raised $30 billion at a valuation of $350 billion. That figure was, at the time, the largest private-market valuation in the company’s history. Three months later, Bloomberg reported this week that Anthropic is in early talks to raise at least another $30 billion — this time at a pre-money valuation exceeding $900 billion. The round may ultimately reach $50 billion, and that investor demand is so intense that the final valuation could breach the trillion-dollar threshold before the books close.
To put those numbers in context: if Anthropic closes this round at $900 billion, it will surpass OpenAI — valued at $852 billion in its March funding round — as the most valuable private AI company in the world. A company founded in 2021 by a group of former OpenAI researchers, with a stated mission to build AI safely, would be worth more than all but a handful of publicly listed corporations on earth. The round is expected to close before the end of May.
What the Revenue Justifies — and What It Does Not
The valuation has not materialised from speculation alone. Anthropic’s annualised revenue stood at approximately $9 billion at the close of 2025. By May 2026, that figure has surged to more than $44 billion — a fivefold expansion in under six months, driven overwhelmingly by enterprise adoption of its Claude and Claude Code products. Eight of the Fortune 10 companies are now stable clients. Over 1,000 enterprise customers spend at least $1 million annually. Gross margins have climbed from 38 per cent a year ago to more than 70 per cent today — a trajectory that, if sustained, addresses the most persistent concern about AI companies: that the cost of generating revenue would perpetually outpace the revenue itself.
This growth rate benefited largely from the rapid enterprise-level implementation of developer tools, such as Claude Code, with enterprise clients contributing over 80 per cent of revenue. That concentration matters. Enterprise customers are stickier, more predictable, and less susceptible to the whims of consumer sentiment than mass-market users. They also, critically, embed the tools they adopt into operational workflows — which makes switching both costly and disruptive. Anthropic’s customer base is, in other words, increasingly structurally locked in.
Even so, the arithmetic of the valuation demands scrutiny. A $900 billion pre-money figure implies a revenue multiple of approximately 20 times forward annualised earnings — a level that assigns enormous weight to growth projections extending well beyond any visible planning horizon. The AI industry has repeatedly demonstrated the capacity to grow faster than conservative forecasters predicted. It has also demonstrated a consistent tendency to consume capital at a rate that would embarrass any ordinary business model.
The Compute Imperative
The stated purpose of this round — and this is a detail that deserves more attention than it typically receives in the breathless coverage of nine-digit valuations — is not product development or market expansion in the conventional sense. Anthropic is raising what is likely to be its last private round before going public, primarily to fund its massive computing needs.
The company has struck a deal with SpaceX to access 220,000 GPUs through the Colossus supercomputing cluster. It has committed $100 billion to Amazon Web Services over the next decade for Trainium chip access and data centre capacity. The infrastructure bill for remaining competitive at the frontier of AI development is, in short, extraordinary — and it is recurring. Each new generation of models requires more compute than the last, on timelines that compress faster than the revenue cycles needed to self-fund them. This is the structural reality that makes the private fundraising market for frontier AI companies unlike any that has preceded it: the capital requirements are not diminishing as the technology matures. They are growing.
The IPO Shadow
The capital push comes as the San Francisco-based company explores a potential initial public offering as early as October, with raising significant funds now viewed as essential for securing the massive computing power required to sustain its growth trajectory. That framing — raising private capital now to fund the compute needed to justify a public valuation later — is the implicit logic of the current round. It is a bridge, not an endpoint.
Public market investors appear to favour Anthropic’s enterprise-oriented business model, viewing its path to profitability as clearer and more verifiable than competitors. That assessment is plausible, and the margin improvement documented over the past year lends it credibility. But public markets apply disciplines that private funding rounds do not. Revenue multiples of 20 times are defensible when growth is accelerating, and the competitive moat is widening. They become precarious the moment either of those conditions reverses.
The deeper question this week’s news poses is not whether Anthropic is worth $900 billion today, but whether the infrastructure economics of frontier AI — the perpetual capital hunger, the compute arms race, the concentration of dependency on a small number of cloud providers — can sustain the valuations the private market is currently assigning. The answer to that question will not be found in a term sheet. It will be found in the revenue lines of the companies building on top of AI, and in whether those lines grow fast enough to justify what is now being asked of investors.
That reckoning, for Anthropic and for the industry it increasingly defines, is scheduled for October.
Read Also: OpenAI bets $300 million on Singapore as first international applied AI hub
