Anthropic's $2 Trillion IPO Test: Can AI Growth Become Profit?
IPO
A first public sale of a company's shares.
valuation
The estimated price of the whole company.
operating loss
When a company's main work costs more than it earns.
What happened
Anthropic, the AI company behind Claude, is preparing a possible initial public offering, or IPO. Reports have placed its possible valuation near $2 trillion. That is not a final price.
CNBC, a financial news outlet, reported on October 7 that New Constructs, an independent financial research firm, called the possible offering the “most ridiculous IPO of 2026.” New Constructs estimated Anthropic at about $150 billion. That estimate sits far below the possible $2 trillion figure.
This distinction matters. The reported fact is that one research firm criticized the proposed price. It is not proof that Anthropic has failed, or that the IPO has been canceled.
Background: what an IPO price means
An IPO lets a private company sell shares to public investors. The valuation is the implied price of the whole company. It does not mean the company receives that amount in cash. It also does not guarantee future profits.
AI companies are often valued on expected growth. Investors may pay more today if they expect much larger sales later. The key question is what remains after computing, staff, research, and other costs.
Why the criticism matters
New Constructs says a $2 trillion valuation would require Anthropic to produce roughly twice Nvidia's latest annual profit. That is a condition from the firm's model. It is not a promise from Anthropic.
The report uses very aggressive assumptions. Revenue would grow 1,000% in 2026 and double in 2027. It would then grow 56% each year through 2035. The model also assumes a 10% profit margin. These assumptions show what the valuation demands. They do not predict what will happen.
What the available documents show
The New Constructs report cites about $4.6 billion in 2025 revenue, a $42 billion net loss, and an $8 billion operating loss. It says the net loss included a large accounting charge. That means the figures describe different parts of the business and should not be treated as identical measures.
The report also points to $518 billion in reported cloud, computing, and infrastructure commitments. This describes possible future obligations. It does not mean Anthropic has already paid all of that money.
The story drew strong attention on Hacker News, a technology forum. The thread had 240 points and 213 comments. Those numbers show community interest. They do not prove that New Constructs is right, or that the $2 trillion figure is wrong.
What remains unknown
The final offering date, share count, price, and valuation remain unsettled. Anthropic also must show whether fast sales growth can become lasting profit. It is unclear how long customers will keep spending, how computing costs will change, and how competition will affect prices.
New Constructs' conclusion also depends on its own growth, margin, and cost assumptions. A different set of assumptions could produce a different valuation.
What to watch next
The next useful evidence will be the final public filing and the actual offering price. Readers should compare revenue, operating profit, cash flow, customer concentration, and long-term computing commitments.
The central question is not which headline sounds stronger. It is whether the evidence connects today's sales with durable future profit.
Sources: CNBC, New Constructs, and Hacker News.
Why Anthropic's Possible IPO Has Two Very Different Price Tags
📰 Full story: Anthropic's $2 Trillion IPO Test: Can AI Growth Become Profit?
Anthropic may seek a $2 trillion public value. One research firm says that price looks far too high.
IPO
A first public sale of company shares.
valuation
The estimated price of the whole company.
operating loss
When the main business spends more than it earns.
💡 The gist
- Anthropic may seek a public value near $2 trillion.
- New Constructs estimates its value near $150 billion.
- Hacker News showed interest, not proof of either number.
Anthropic makes Claude, an AI assistant. It may sell company shares publicly. This first sale is called an IPO.
An IPO gives buyers small ownership pieces. Valuation means the estimated price of the entire company. A possible $2 trillion valuation would be enormous. It is not money already sitting in Anthropic's bank account.
CNBC, a financial news outlet, reported New Constructs' criticism. New Constructs is an independent financial research firm. It says the possible price expects too much future growth.
AI services need expensive computers. They also need workers and research. Fast sales growth matters only if enough money remains afterward. That is why sales and profit must be checked separately.
New Constructs says Anthropic would need about twice Nvidia's recent profit. Nvidia is a major chip company. This is a calculation, not a company promise.
The report cites about $4.6 billion in 2025 sales. It also cites a $42 billion net loss. Its operating loss was about $8 billion. A large accounting charge affected the net-loss figure. The report also mentions $518 billion in future cloud and computing commitments.
Hacker News, a technology forum, gave the story 240 points and 213 comments. That shows attention. It does not prove the research firm's estimate.
The final IPO price is still unknown. The timing and share count are also unknown. Readers should watch the final filing, revenue, profit, cash flow, and computing costs.
Sources: CNBC, New Constructs, and Hacker News.
💬 Easy summary: Anthropic may be useful but still overpriced
The comments are divided. The revenue, success-rate, and share-price figures mentioned by commenters are self-reported and independently unverified.
- The discussion centers on a roughly $2T valuation. A useful AI product does not automatically make that price fair.
- Bullish commenters say Anthropic can sell an enterprise service and that open models have not taken over mainstream demand. One commenter reports $60B in revenue for Anthropic versus $140M for OpenRouter, but those figures are unverified.
- They also say frontier AI can make a task go from about 60% reliable to 99% reliable, creating new demand. That performance gap is a commenter-reported claim.
- Skeptics say future training, servers, chips, salaries, and reinvestment can consume more cash than inference earns. Cheap or open models can catch up, and software can switch models, so Anthropic's premium may shrink.
- An Uber stock recovery does not prove its IPO price was a good buy. Some commenters also predict an AI crash, while others note that bubbles can be widely recognized before they burst.
initial digest at 213 comments (revision 1). We fetched 100 comments and sampled 100 across the thread. These are HN users’ reports, not independently verified facts.
Anthropic Has Two Company Price Tags
📰 Full story: Anthropic's $2 Trillion IPO Test: Can AI Growth Become Profit?
One price is huge. Another price is much smaller.
IPO
A way to sell small pieces of a company to many people.
Hacker News
A place where people share technology news.
A little story
Anthropic, the company that makes Claude, may sell tiny shares. This sale is called an IPO. An IPO lets many people own small pieces.
People are discussing a $2 trillion company price. New Constructs, a number-checking company, says $150 billion looks closer. Those are two very different prices.
Anthropic made $4.6 billion in sales in 2025. It also had an $8 billion operating loss. That means its main work cost more than it earned.
New Constructs says Anthropic needs huge future profits. It says those profits must reach twice Nvidia's recent profits. Nvidia is a chip company. These are guesses from a money calculation. They are not finished facts.
Hacker News is a place where people share technology news. This story got 240 points and 213 comments. That shows attention. It does not show which price is true.
The real IPO price is not set yet. We must watch whether Anthropic can earn money later.
Sources: CNBC, New Constructs, and Hacker News.
💬 For a five-year-old: a story about an AI company's stock
An AI can be useful, but the company may still spend too much money. The numbers are claims made by commenters, not checked facts.
- People discuss a price of about $2T for Anthropic. But useful does not always mean worth that much.
- One commenter says Anthropic makes about $60B and another AI service makes $140M. Another says a strong AI can make a job work about 60% to 99% of the time. These are the commenters' own claims.
- The company must keep paying to teach the AI and run many computers. Cheaper public models may do the job too, so the company may keep less money.
- A stock can fall after an IPO and rise later. The Uber example and the AI-bubble predictions do not prove that the first price was good. Useful technology and a good investment are different.
initial digest at 213 comments (revision 1). We fetched 100 comments and sampled 100 across the thread. These are HN users’ reports, not independently verified facts.
💬 Anthropic IPO: Useful AI Is Not the Same as a Justified Stock Price
The comments split over whether Anthropic can turn useful frontier AI into durable cash returns. They discuss a roughly $2 trillion valuation, but all revenue, performance, usage, and share-price figures below are commenters' self-reported claims, not independently verified.
initial digest at 213 comments (revision 1). We fetched 100 comments and sampled 100 across the thread. These are HN users’ reports, not independently verified facts.