Summary of the Key Points in Plain Language
The biggest news in the global capital markets recently is that Anthropic, the company behind the renowned large-scale AI model Claude, is preparing for what could be the largest IPO in human history. It is expected to list on the U.S. stock market by the end of September, aiming to raise over $100 billion and target a valuation of $2 trillion, which would surpass the previous record set by SpaceX in June with a raise of $8.57 billion and a valuation of $1.77 trillion.
Anthropic has almost completed all its pre-listing preparations: more than a dozen top global banks have agreed to provide it with a preliminary loan of $15 billion. Its financial performance is also impressive—revenue from paid U.S. customers has surpassed that of its rival OpenAI, and the company has seen a sevenfold increase in revenue within a year, with it already turning a profit. However, it faces several challenges. SpaceX, the previous “super IPO star,” saw its stock price drop by a third within half a month after listing, dampening the overall enthusiasm for such large-scale offerings. Additionally, issues such as insufficient computing power, competition from open-source models, and an excessively high valuation pose significant risks to this “century IPO.” The entire market is closely watching to see whether Anthropic will succeed or fail.
---
Detailed Breakdown of the Key Points
1. More than a Dozen Top Banks Competing for a $15 Billion Loan: A Way to Secure an IPO Ticket
Many may wonder why Anthropic, which is already in the process of listing and doesn’t seem to lack funds, would need to borrow $15 billion from banks. Top investment banks like Morgan Stanley, Goldman Sachs, and JPMorgan are all eager to participate, with the largest bank offering $1.25 billion and others contributing at least $750 million each. This is a common practice for large-scale U.S. IPOs. For example, SpaceX increased its credit limit from $1.5 billion to $5 billion just one month before its listing. This loan serves as “bridge capital,” allowing Anthropic to purchase chips, build data centers, and acquire computing power without waiting for the IPO funds to be received, thus not delaying its expansion plans. Moreover, banks that participate in these loans gain the opportunity to become the lead underwriters for the IPO, potentially earning tens of millions or even hundreds of millions in commission. JPMorgan, for instance, rose to the top of the U.S. stock issuance rankings by securing the SpaceX deal, making this a profitable investment.
2. The $2 Trillion Valuation Is Not Just Bluff: Its Performance Outpaces Its Competitors
The $2 trillion valuation may seem ambitious, but looking at Anthropic’s financial records, its growth rate has surpassed all previous records in the tech industry:
- Rapid Profit Growth: Its annual revenue was only $9 billion in 2025, but by July 2026, it expected to earn $65 billion, a sevenfold increase in just seven months, with $11.5 billion already earned in the second quarter alone.
- Market Share Lead: 34.4% of U.S. companies are now willing to pay for its services, surpassing OpenAI’s 32.3% and making it the leading enterprise-level AI model in the U.S.
- Profitability: While other AI companies were losing money, Anthropic has achieved a gross margin of 70%-85% in its inference services and turned a profit in the second quarter. Its latest model has doubled in performance while reducing usage costs by up to 45%, enhancing customer loyalty.
Its valuation reflects its strong performance. The valuation increased from $183 billion in November last year to $380 billion in February and then to $965 billion in May, setting expectations for further growth. This is not just based on AI hype but on solid financial fundamentals.
3. SpaceX’s Setback Has Dropped the Hype for Anthropic’s IPO
Anthropic’s IPO timing is particularly challenging. SpaceX, which was once hailed as a “century IPO” example, saw its stock price drop by 33% within two months of listing, with its market value shrinking by over $1 trillion. This setback has dampened market enthusiasm for similar offerings. Additionally, the high cost of computing power and competition from open-source models create uncertainties for Anthropic’s IPO.
4. Three Major Challenges Ahead
To achieve its $2 trillion valuation, Anthropic must overcome three significant hurdles:
- Insufficient Computing Power: It currently spends $12.5 billion per month on computing power from SpaceX and has a five-year, $200 billion cloud service contract with Google. However, 71% of Americans oppose the construction of new AI data centers, and policies in Florida and Pennsylvania have restricted such projects. If computing power expansion lags, even high revenue growth targets will not be achievable.
- Competition from Open-Source Models: Free or low-cost open-source models can meet the needs of many small and medium-sized businesses, potentially eroding Anthropic’s market share and profits. If these models gain popularity, Anthropic’s margins could be significantly impacted.
- Excessively High Valuation: A valuation of $2 trillion implies a high growth rate, but it has not yet generated substantial net profits. Any slowdown in revenue growth could lead to a bubble bursting in its valuation. The current valuation is largely based on future growth expectations.
In essence, Anthropic’s IPO is a test for the entire AI industry. If it succeeds, other AI companies’ valuations could soar; if it fails, it could cool the industry’s funding momentum. The entire tech community is eagerly awaiting its listing on September 30th to see how this story plays out.