What You'll Learn
- The exact details of Anthropic's confidential June 1, 2026 IPO filing, including the $965 billion valuation, the $65 billion Series H, and the lead underwriters
- How Anthropic grew from $1 billion to $30 billion in annualized revenue in 14 months, and why that growth rate has no peer in the software industry
- Why 8 of the Fortune 10 now pay Anthropic for Claude and what the $100 billion AWS infrastructure deal means for the public-market story
- How the Anthropic, OpenAI, and SpaceX IPOs together could demand $200 billion from public markets and the second-order stocks that move when Anthropic lists
Anthropic's confidential IPO filing on June 1, 2026 sent a tremor through Wall Street that ended any remaining debate about whether AI is a real industry or a venture capital fever dream. The Claude maker, founded five years ago by seven former OpenAI researchers, disclosed in a routine submission to the Securities and Exchange Commission that it intends to become a publicly traded company, a step that arrives less than a week after the firm closed the largest private funding round in the history of artificial intelligence. The $65 billion Series H, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, valued Anthropic at $965 billion post-money, a number that briefly surpassed OpenAI's most recent private mark and put the San Francisco company within striking distance of a $1 trillion listing.
But the valuation, eye-watering as it is, is only the headline. The S-1 also signals a profound shift in how AI labs are positioning for the public markets. Anthropic is the first of the three foundational model companies, OpenAI, xAI, and Anthropic, to formally pull the trigger. And unlike the SpaceX S-1, which rests on launch revenue and Starlink subscription dollars, Anthropic's filing is anchored to enterprise software economics. With $30 billion in annualized run-rate revenue, eight of the Fortune 10 on the customer list, and a ten-year, $100 billion cloud infrastructure commitment from Amazon, Anthropic is asking Wall Street to value it as the operating system of the agentic economy, not just a model provider.
This article walks through every fact in Anthropic's filing that investors and operators need to know, the strategic logic behind the timing, and how the Anthropic IPO slots into the most ambitious year of public debuts the markets have ever seen.
Anthropic Files Confidentially: The Day Wall Street's AI Obsession Went Public
Anthropic's confidential submission landed at the SEC on Monday, June 1, 2026, according to reporting from The New York Times, Reuters, and The Wall Street Journal, all of which broke the story within hours of one another. A confidential filing allows the company to share its registration statement with regulators privately, withhold sensitive financial details from the public eye during the SEC's review process, and only make the full S-1 public a few weeks before the roadshow begins. The structure has been used by nearly every major tech IPO of the last decade, including Facebook, Airbnb, and Snowflake, precisely because it gives issuers more control over the narrative arc.
The filing itself, however, is not the only market-moving disclosure. Anthropic's Series H closed on May 28, 2026, just four days before the S-1 went to the SEC. Investors including Altimeter Capital, Dragoneer Investment Group, Greenoaks Capital, and Sequoia Capital committed $65 billion of new capital, lifting the company's post-money valuation to $965 billion. The Wall Street Journal reported that the round was priced at a premium to OpenAI's $852 billion tender offer earlier in the spring, formally making Anthropic the most valuable private AI lab in the world.
The share-price implications of a $965 billion private round, layered on top of a confidential IPO, are almost without precedent in modern market history. For context, the entire 2025 US IPO market raised roughly $45 billion, according to Renaissance Capital. The Anthropic listing alone, if it prices near its last private mark, would be more than 20 times the total proceeds raised by every US IPO last year combined. That is the scale of the event Wall Street is preparing for.
But the real question, the one every Goldman Sachs partner and every Fidelity portfolio manager spent the evening of June 1 working through, is not whether the IPO will be large. It is whether a company that raised $65 billion in private capital just four days before filing can credibly be sold to public investors at a premium without leaving significant paper gains on the table for late-stage venture funds. Anthropic's underwriters, who have not yet been named, will have to thread that needle carefully.
Inside the $965 Billion Valuation: How Anthropic Built the Largest AI Funding Round in History
The Series H was, in raw dollars, the largest single private financing in the history of artificial intelligence and one of the ten largest private financings ever recorded in any industry. To put the $65 billion round in context, the previous high water mark was OpenAI's $40 billion round in March 2025 at a $300 billion valuation. Anthropic has now raised $65 billion at 3.2 times that valuation in just 14 months.
The investor syndicate was deliberately built for the long public markets game, not for the late-stage private churn that defined AI in 2024. Altimeter Capital, the lead investor, is a Boston-based growth fund that has held Anthropic shares since 2024 and is widely expected to anchor the IPO order book. Dragoneer and Greenoaks, both long-duration growth investors, have written public letters arguing that AI infrastructure will resemble a utility-scale buildout, with the same compounding economics. Sequoia Capital, Anthropic's earliest institutional backer, doubled its position in the round.
The price per share implied by the $965 billion post-money mark has not been disclosed, but secondary market trades in late May put Anthropic shares at roughly a 30% premium to the last primary round, meaning a $1.2 trillion implied mark for any new investor buying into the IPO. That secondary activity is significant because it represents a real-money vote of confidence from hedge funds and family offices that did not get allocation in the primary Series H. When the S-1 goes public, those secondary prices will be the most important comparable for the IPO valuation range.
What the Series H was really financing, however, is the buildout of compute capacity. Anthropic has now committed more than $100 billion of forward cloud spending to Amazon Web Services over the next decade, with capacity reservations of up to 5 gigawatts of Trainium and Nvidia GPU clusters. The Series H, combined with a parallel $25 billion equity investment from Amazon announced in April 2026, gives Anthropic a runway to scale Claude inference at a level that no peer outside of OpenAI and Google DeepMind can match (see our Computex 2026 coverage for how the AI silicon race is reshaping this dynamic). It is the capital base of a hyperscaler, and that is the economic moat Anthropic will be selling to public investors.
From $1 Billion to $30 Billion: Anthropic's Revenue Rocket vs Every Other AI Lab
The single most important number in Anthropic's filing, and the one that will most directly drive the IPO valuation, is annualized run-rate revenue. As of May 2026, the company disclosed internally to its bankers that its run-rate is approximately $30 billion, according to the New York Times and Reuters. That is up from $9 billion at the end of 2025, up from $4 billion in July 2025, and up from $1 billion in December 2024. The company has grown annualized revenue roughly 30 times in 14 months.
To benchmark that against the rest of the AI sector, OpenAI's most recently disclosed annualized run-rate revenue is approximately $13 billion as of late 2025, with the company guiding to roughly $20 billion by the end of 2026. Anthropic, by contrast, will likely cross $40 billion in run-rate revenue before the end of 2026, meaning the Claude maker will book more in 12 months than OpenAI's best estimate of 2027. Claude Code, Anthropic's agentic coding product, has independently grown to a $2.5 billion run-rate on its own, with usage climbing roughly 5x between the Series F and Series H rounds.
The breakdown of that $30 billion matters even more than the headline. More than half of Anthropic's revenue now comes from enterprise API consumption rather than consumer subscription products, according to Business of Apps and Anthropic's own Series G disclosure. The number of customers spending more than $1 million per year on the Claude API has crossed 500, up from a dozen in early 2024. The number spending more than $100,000 per year has grown 7x in the last twelve months. The 80/20 of AI revenue, the one that determines whether a public company is a software-as-a-service business or a consumer-internet business, is decisively enterprise for Anthropic.
This is the financial signature that has historically earned software companies the highest public-market multiples. Microsoft, Salesforce, ServiceNow, and Snowflake have all traded at 15x to 30x forward revenue during periods of similar growth. If Anthropic prices the IPO at even a conservative 25x forward revenue on a $40 billion 2026 exit run-rate, the implied market cap would be $1 trillion, in line with the secondary market already trading at today. The math, for once, is on the company's side.
The $100 Billion AWS Bet: How Amazon Locked In the Decade That Powers Claude
The most consequential agreement in the Anthropic S-1, the one that will define the company's cost structure and competitive moat for the next decade, is the $100 billion cloud infrastructure commitment to Amazon Web Services. The deal was announced in expanded form in April 2026, building on an earlier 2023 agreement that already had Anthropic as one of AWS's largest customers. Under the new terms, Anthropic has committed to spend more than $100 billion on AWS technologies over the next ten years, including up to 5 gigawatts of Trainium 3 and Nvidia Blackwell capacity for training and inference. In return, Amazon has invested an additional $25 billion into Anthropic's equity, bringing Amazon's cumulative investment to roughly $25 billion, of which more than $60 billion is now paper-profit on the $965 billion mark.
The strategic logic of the deal is two-sided. Anthropic gets a guaranteed compute runway that no peer can match without comparable capital. AWS gets the most credible independent AI lab locked into its silicon and its cloud for a generation. For public-market investors, the deal does three important things. First, it removes compute cost as a near-term risk factor. Second, it gives Amazon's AWS division a direct line into the upside of every Claude API call. Third, it makes it almost impossible for a hyperscaler like Microsoft Azure or Google Cloud to acquire Anthropic outright, because Amazon's economic position would make any competing bid uneconomic.
The 5 gigawatts of reserved capacity is itself a stunning number. Total global data center capacity in 2024 was approximately 90 gigawatts across all hyperscalers, according to the International Energy Agency. Anthropic alone has reserved nearly 6% of all global cloud capacity for a single decade, a level of resource concentration that no other AI lab has achieved. The Claude maker, in effect, has bought itself a sovereign-grade compute footprint before going public. That is what the $65 billion Series H was really financing.
Anthropic vs OpenAI vs SpaceX: The Most Concentrated IPO Wave in Market History
The Anthropic filing is the third leg of a 2026 IPO wave that will, in absolute dollar terms, be the largest concentration of new public equity the markets have ever seen. The SpaceX S-1, filed in March 2026, is targeting a market cap in the $2 trillion range. OpenAI is widely expected to file its own confidential S-1 within weeks, with a target listing date in late 2026 and a valuation range of $852 billion to $1 trillion. The three filings together could demand upwards of $200 billion of public-market capital, a sum equal to roughly 60% of the entire US money-market fund industry's total assets.
The comparison is useful because the three companies, despite all being flagship AI-era listings, are fundamentally different businesses. SpaceX is a capital-intensive industrial company with launch services and Starlink subscription revenue. OpenAI is a hybrid consumer subscription and developer API business with a heavier consumer tilt. Anthropic is the most enterprise-concentrated of the three, with the highest gross margin profile, the most repeatable revenue, and the longest-dated compute commitments. For a public-market investor choosing exposure to the AI thematic, the three are not substitutes but complements.
There are also second-order beneficiaries that the Anthropic filing has already moved. Zoom Video Communications holds approximately $1.27 billion of Anthropic equity on its balance sheet, and the stock rose 11% on June 1, 2026, the day the S-1 became public. Salesforce, an early Anthropic investor through Salesforce Ventures, also saw a sharp move higher, with TipRanks flagging it as a top AI IPO play. These movements matter because they confirm that the public markets are treating Anthropic as a systemically important AI asset, not just a private-equity trophy.
The risk for retail investors, and the one that the New York Times specifically flagged in its June 1 coverage, is that IPOs priced at $500 billion or higher tend to underperform in the 12 months following listing. The 2018 to 2019 mega-IPOs of Uber, Lyft, and Slack all priced at peak private valuations and then traded down significantly. Anthropic's bankers will need to find a price that gives the company credit for its growth but also leaves enough upside to compensate public investors for the risk of holding a 30-bagger that has already priced in 50% of its growth.
The Fortune 500 Tsunami: Why 8 of the Top 10 Companies Now Pay Anthropic
The most underappreciated line item in the Anthropic S-1, and the one that will most directly support the post-IPO multiple, is customer concentration. According to independent reporting by The Information, Fortune, and Anthropic's own Series G press releases, eight of the Fortune 10 now pay Anthropic for Claude API access. That includes every major US bank, the two largest credit card networks, three of the four largest professional services firms, and both of the largest cloud platforms outside of Amazon itself. The remaining two of the Fortune 10 are, by public process, also Claude customers in some form.
The number of customers spending more than $1 million per year on Claude has crossed 500, up from a dozen in early 2024, according to Anthropic's Series G announcement and Medium analysis of the customer roster. The number spending more than $100,000 per year has grown 7x in twelve months. These are not pilot customers or proof-of-concept deployments. They are production workloads running core business processes. Claude is not a research toy for the Fortune 500. It is the new system of record for legal review, financial modeling, customer support routing, and code generation, in line with how Claude Opus 4.8's dynamic workflows have moved from research preview to enterprise deployment.
The enterprise wedge matters for three reasons. First, enterprise customers have extremely high switching costs once a model is integrated into a business process. Switching a customer support bot from Claude to GPT or Gemini is a 6 to 12 month engineering project, not a config change. Second, enterprise contracts are typically multi-year with annual price escalators, giving Anthropic visible forward revenue that consumer subscription businesses cannot match. Third, enterprise gross margins are 70% to 80%, compared with 40% to 50% for consumer API businesses, because enterprise customers buy committed capacity that smooths out the cost-of-revenue curve.
The Claude Code agentic product, the second pillar of the enterprise story, has now grown to a $2.5 billion annualized run-rate on its own, according to Anthropic's Series G announcement. Claude Code's coding assistant market share in the United States is estimated at 70% by some analysts, ahead of GitHub Copilot, Cursor, and direct OpenAI access. With 8 of the Fortune 10 as customers, $500 million or more in average annual spend per top-tier enterprise, and a 70% share of the US coding market, the company has assembled an enterprise footprint that took Snowflake a decade and Salesforce two decades to build.
The Second-Order Trade: Stocks That Move When Anthropic Lists
Beyond the direct trade on Anthropic's IPO, there is a parallel and arguably more interesting trade for US investors: the basket of public companies that will reprice when Anthropic lists. Three categories matter.
The first is the public-cloud-company exposure. Amazon's $25 billion cumulative equity investment in Anthropic is now worth more than $60 billion on paper at the $965 billion mark, a 140% unrealized gain in less than 24 months. Microsoft's exposure to OpenAI is well known and broadly priced in, but the Anthropic-Amazon relationship is materially underpriced by the market. If Anthropic prices the IPO at $1 trillion, the implied value of Amazon's stake alone would add roughly $40 billion to Amazon's market cap on a mark-to-market basis, a 1% lift on a $2 trillion company. The AWS business also benefits from the $100 billion infrastructure commitment, which is essentially a forward revenue contract from a single customer.
The second category is the customer-stake basket. Salesforce Ventures invested in Anthropic through its corporate venture arm, and Salesforce's stock has historically moved 3% to 5% on AI partnership news; see our Salesforce Q1 FY27 earnings recap for context on the Agentforce monetization. Zoom's $1.27 billion Anthropic stake is even more concentrated, and the stock's 11% move on June 1, 2026, the day of the filing, demonstrated exactly how much leverage retail investors have to a single AI event. Other companies with disclosed Anthropic exposure include Google (through its cloud partnership), NVIDIA (as a compute supplier), and several smaller enterprise software names.
The third category is the AI infrastructure basket, the names that supply the picks-and-shovels to the entire AI buildout. NVIDIA, Broadcom, AMD, and the HBM memory trio of SK Hynix, Micron, and Samsung all benefit from the $100 billion AWS-Anthropic compute commitment. Anthropic alone is now a top-five customer for several of these companies. If the IPO prices well and Anthropic accelerates its infrastructure spend in 2027, these picks-and-shovels stocks will reprice as the embedded AI capex tailwind becomes more visible to public investors.
The takeaway for US allocators is straightforward. The Anthropic IPO is not just a single trade. It is a system event that re-prices the entire AI value chain, and the best risk-adjusted exposure is often in the established public-market names that have been the second-order beneficiaries all along.
Conclusion
Anthropic's confidential IPO filing on June 1, 2026 marks the moment the AI sector stopped being a private-capital story and became a public-market story. With $30 billion in annualized run-rate revenue, a $100 billion AWS commitment, 8 of the Fortune 10 as customers, and a $965 billion valuation, Anthropic has assembled the most enterprise-grade AI business the markets have ever seen. The S-1 is the largest of three AI-native public offerings expected in 2026, alongside SpaceX and OpenAI, and the three together will define a new era for the technology sector and the IPO market. For retail investors, the listing will be the first real opportunity to buy a piece of the foundational AI economy at scale. For institutional allocators, the pricing will set the comp for every AI business that comes to market for the next five years. Anthropic is no longer the insurgent. It is the establishment, and it is coming to a brokerage account near you.
Anthropic and OpenAI Launch Wall Street AI Joint Ventures: The $15.5 Billion Race to Become Enterprise AI's Operating System
External Reference: Anthropic Series H Announcement (Official) — Full details on the $65 billion round, lead investors, and the $100 billion AWS infrastructure commitment.
Last Updated: June 02, 2026 | Source: Anthropic, Inc. (anthropic.com); SEC EDGAR Filings; The New York Times; Reuters; The Wall Street Journal; Fortune; Bloomberg; CNBC