
A Valuation Set by the Players, Not the Market
Picture a private card game with four players: Amazon (AMZN), Google (GOOGL), Nvidia (NVDA), and Anthropic. They pass chips around, and each pass raises the value. In March 2025, Anthropic was valued at $61 billion. By May 2026, that figure was $965 billion. The IPO now prices the company at $2 trillion.
The core problem: no outside buyer has ever paid these prices. The S1 filing shows no public trades that set a real market price. Amazon, Google, and Nvidia all invested in Anthropic and are tied together. Amazon (AMZN) and Google (GOOGL) own part of the company and also sell it computing power. Almost half of last year's sales came through those two. When the price rises, these investors book the profit. Amazon (AMZN) booked about $53 billion in profit in the first quarter from re-valuing its Anthropic stake. Nobody outside this group paid that price.
Now 22 banks are asking the public for $2 trillion, double the last mark. The outsiders joining the table are being asked to pay it. Whether that price can hold is the big open question.
How the $2 Trillion Was Calculated
The valuation works backward from a guess. They took projected 2028 revenue of about $195 billion and multiplied it by 10, which gives $2 trillion. Use today's sales of $65 billion instead, and a 10x multiple gives a valuation of $650 billion. Same company, same day, only the input number changes. The gap between $650 billion and $2 trillion is $1.35 trillion of price resting on a guess about where revenue lands two years out.
The Demand Is Real; the Numbers Are Growing Fast
Sales are climbing hard. Revenue was $386 million in 2024, $4.6 billion in 2025, and $16 billion in the first six months of this year. About 12,000 customers pay over $100,000 a year, but almost half of that came from Amazon (AMZN) and Google (GOOGL). The company posted its first real profit last quarter: $974 million.
Concentration Risk and the Industry Funding Gap
More diversified revenue is always better. Google (GOOGL) and Amazon (AMZN) are very large companies and give some assurance of payment. But a valuation forecast must weigh liabilities, not just the value a company brings.
A broader gap hangs over the whole sector. Gartner issued a report stating that by 2028 the AI industry would need about $8.6 trillion to cover all its capital spending, while projecting industry revenue of only $3.6 trillion. That leaves roughly a $5 trillion gap, and no one knows how to close it.
Competition and the Free-Model Threat
Growth so far came when Anthropic had only one main rival. Now it faces OpenAI, Google (GOOGL), and other AI names at home and abroad. The question is whether the Claude models are different enough to hold the projected growth rate.
People clearly want AI, but wanting the price is another matter. Frontier model makers like OpenAI and Anthropic carry huge infrastructure costs. Meanwhile, Chinese models sit only three to six months behind and are nearly at the same level, and they offer their product almost for free. A competitor giving away roughly the same product for free is a serious worry for any company trying to charge for it.


