In May 2021, Anthropic raised a $124 million Series A at an $845 million valuation. Five years later, the same company filed confidentially for an IPO at a $965 billion mark, surpassing OpenAI, and quietly delivering the largest startup wealth event of the past decade. The story of how an AI safety lab took the throne from the company it broke away from is, before anything else, a story about capital.

On June 1, 2026, Anthropic confidentially filed a draft Form S-1 with the Securities and Exchange Commission, the first formal step toward a public listing that could occur as early as fall 2026 (Mickle, 2026; Los Angeles Times, 2026). The filing arrived four days after Anthropic closed a $65 billion Series H at a post-money valuation of $965 billion, surpassing OpenAI’s $852 billion private mark and making Anthropic the most valuable private AI company in the world (CNBC, 2026).

The bare facts are striking enough. The deeper story is in the multiples. The investors who put $124 million into Anthropic’s Series A in May 2021 now hold paper positions worth, on the current valuation, over a thousand times their original cost basis. The story of how that happened, who profited most, and what it signals about the next two years of AI capital markets is the story this article tells.

The Numbers Behind the Throne

The chart below traces every named Anthropic funding round from Series A in May 2021 through the Series H announced on May 28, 2026, and the implied return multiple at each round for an investor who entered at the Series A price.

Three patterns are visible.

The first is that Anthropic’s valuation moved in a relatively orderly fashion through 2024. The company stayed in the $4-billion-to-$18-billion range from Series B in 2022 through Series E in February 2024 (Dealroom, 2026; Anthropic, 2021). This was the post-ChatGPT period when investor enthusiasm for AI was high but Anthropic was still considered a clear second to OpenAI on capability and brand recognition.

The second is the inflection. The Amazon $4 billion investment in November 2024 pushed Anthropic to a $40 billion valuation. Series F in March 2025 reached $61.5 billion. Series G in February 2026, led by ICONIQ Capital with participation from Coatue, D. E. Shaw, Dragoneer, Founders Fund, GIC, and MGX, plus participation from BlackRock, Blackstone, Microsoft, Nvidia, Sequoia, and Goldman Sachs Alternatives, hit $380 billion (StartupIntros, 2026). And the Series H closed three months later at $965 billion (Pequeño, 2026).

The third is the magnitude of early-investor returns. The Series A round was led by Jaan Tallinn, James McClave, Dustin Moskovitz, Eric Schmidt, and the Center for Emerging Risk Research, with a total raise of $124 million at an $845 million post-money valuation (Anthropic, 2021). At the current $965 billion mark, that 2021 entry has, on paper, multiplied by approximately 1,142 times. A $1 million Series A check is currently valued at roughly $1.14 billion.

Who Wrote the Series A Check, and Who Wins

The story of who profits most from Anthropic’s IPO begins with the names on the Series A round. They are not the names that dominate AI venture capital headlines today.

Jaan Tallinn, the Estonian co-founder of Skype and Kazaa, has spent the past decade investing in AI safety projects through the Center for Emerging Risk Research (formerly the Survival and Flourishing Fund). Tallinn was a personal lead investor in the Anthropic Series A and, separately, an early supporter of effective altruism and existential-risk research that informed Anthropic’s founding thesis (Anthropic Series A photo, 2021).

Dustin Moskovitz, the Facebook co-founder and Asana CEO, has donated billions through his Open Philanthropy Project, which has funded a substantial share of the world’s AI safety research. Moskovitz’s personal investment in the Series A is consistent with his broader thesis that aligned AI development is the most consequential cause-area he can capitalize.

Eric Schmidt, the former Google CEO and chairman, is the most institutionally connected of the Series A backers. Schmidt’s involvement provided what observers at the time described as social proof at a moment when most of the world’s AI capital was flowing into OpenAI.

James McClave, founder of MX Technologies, was a less-publicized but significant Series A check. McClave’s name appears alongside Tallinn, Moskovitz, and Schmidt in the original Anthropic announcement.

The Series A investors and the Center for Emerging Risk Research together put $124 million into Anthropic at a moment when, in the words of one investor who passed at the time, “OpenAI was the only conversation in San Francisco.” Five years later, that pool of $124 million, if held intact through every subsequent round without dilution beyond the publicly reported employee option pool, would be valued at approximately $14 billion on paper at the $965 billion mark.

The largest single beneficiary of Anthropic’s valuation arc is not a Series A investor. It is Google. According to legal filings in the Department of Justice antitrust case against Google, Alphabet holds a 14 percent stake in Anthropic, accumulated through more than $3 billion in direct investment and a $40 billion commitment announced in April 2026, of which the first $10 billion was paid in cash (Mickle, 2025; Reddit summary of Bloomberg reporting, 2026). At the current valuation, Google’s 14 percent position represents approximately $135 billion in implied value. That is a real, if illiquid, paper gain that ranks among the most consequential single-company tech-investment outcomes in history.

Amazon is the second-largest beneficiary on absolute dollars. After its initial $4 billion commitment in 2023, Amazon committed an additional $4 billion in November 2024 and then a further $5 billion in early 2026 (Capoot, 2025). Total Amazon investment, by the most conservative public reporting, is at least $13 billion. Amazon does not disclose its precise equity stake, but industry observers estimate a 13 to 16 percent position, which would imply approximately $125 billion to $155 billion in current paper value.

The collective implication is that the IPO, when it occurs, will be the largest single-day wealth event for venture capital and corporate strategic investors in the history of the technology industry. Anthropic has raised $74.9 billion across 15 funding rounds through Series H, which is more capital than most publicly traded technology companies have ever raised across their entire histories (StartupIntros, 2026).

The Three-Way IPO Race

Anthropic’s filing places it in a remarkable cohort. SpaceX is expected to price its IPO in the coming days at a targeted valuation of approximately $1.8 trillion. OpenAI has been preparing its own public listing with reported timing for fall 2026 at a private mark of $852 billion. Anthropic now joins both at the largest scale of any IPO cohort in modern memory (Yahoo Finance, 2026; Mickle, 2026).

Combined private valuation of the three: approximately $3.6 trillion. PitchBook noted in May 2026 that the three companies, if they list within twelve months, could match a decade of US venture-capital-backed IPO proceeds in a single year. That comparison is not rhetorical. Total US IPO proceeds in 2024 were approximately $40 billion. Anthropic’s Series H alone raised $65 billion in private capital.

The economic implication is concrete. The public markets are about to absorb three companies that have, between them, no track record of profitability, no public earnings history, and no comparable peer group. Investors who have spent the past 18 months watching their portfolios appreciate on AI exposure will be asked, simultaneously, whether they believe Anthropic and OpenAI deserve $1 trillion-plus public valuations and whether SpaceX deserves a $1.8 trillion mark.

The most credible bearish argument against the trio is straightforward. Anthropic at $965 billion currently trades at approximately 69 times its $14 billion annualized revenue. OpenAI at $852 billion trades at 34 times its $25 billion ARR. By contrast, Nvidia, the most successful technology company of the AI era, trades at approximately 25 times forward revenue at all-time-high earnings. Either the AI lab multiples will compress sharply when these companies hit public markets, or the broader technology multiple framework that has prevailed since 2010 needs to be revised. Both outcomes are possible. They cannot both be true simultaneously.

The most credible bullish argument is also straightforward. If on-device AI inference and agentic computing capture the share of consumer and enterprise software that Microsoft, Nvidia, and the AI labs are betting they will capture, the total addressable market is large enough that current private valuations may, in retrospect, look conservative. Nvidia’s market capitalization has tripled in three years on exactly that bet, and remained credible against most quantitative valuation frameworks while doing so.

How Anthropic Took the Throne

The valuation crossover with OpenAI did not happen by accident. Three economic forces produced it.

The first was enterprise concentration. Anthropic disclosed in its Series G announcement that more than 1,000 companies were spending more than $1 million each on Anthropic services, a level of enterprise spend concentration that produces higher gross margins and lower customer churn than OpenAI’s predominantly consumer-subscription revenue base (Anthropic, 2026). For institutional investors valuing AI labs against software-as-a-service comparable companies, a dollar of enterprise ARR is worth meaningfully more than a dollar of consumer subscription ARR.

The second was the revenue trajectory itself. Anthropic’s annualized revenue went from approximately $1 billion in December 2024 to $4 billion by mid-2025, $9 billion by year-end 2025, and $14 billion by February 2026 (SaaStr, 2026). That is the steepest enterprise SaaS revenue curve ever publicly recorded. Investors paying a premium to that trajectory are pricing not the current revenue but a continuation of the curve for another 12 to 24 months.

The third was governance stability. Anthropic still has all seven of its founding executives. OpenAI has experienced the brief 2023 ouster of Sam Altman, the resignation of co-founder and chief scientist Ilya Sutskever, the departures of Andrej Karpathy and John Schulman (who joined Anthropic in 2024), and ongoing reorganization of its non-profit governance structure. For investors valuing companies on a 10-year horizon, leadership continuity is a material premium.

Dario Amodei, Anthropic’s co-founder and CEO, summarized the contrast at a Bloomberg event in 2024: “We have 7 cofounders. Three and a half years later, we’re all still at the company.” That single line, repeated through every Anthropic fundraising pitch since, has become one of the most quoted competitive comparisons in technology venture capital.

What the IPO Will Reveal

An S-1 filing, even a confidential draft, will eventually require Anthropic to disclose information that the private rounds have not.

The first revelation will be the gross margin. Anthropic has signaled that approximately 50 percent of its annual revenue is spent on cloud compute, primarily through its Amazon Web Services and Google Cloud commitments. Whether the company’s gross margin trajectory is improving as it scales, or whether the compute cost burden is structural, will be one of the most-watched disclosures in the filing.

The second will be the customer concentration. Anthropic has not disclosed how concentrated its $14 billion ARR is across its top 10 customers. If a small number of large enterprise customers, particularly Amazon, Salesforce, Snowflake, and a handful of consulting firms, account for 30 to 50 percent of total revenue, that concentration is both a strength (durable enterprise relationships) and a vulnerability (loss of any one customer could materially impair the revenue curve).

The third will be the financial relationship with its corporate investors. Anthropic’s commercial arrangements with Amazon and Google involve compute commitments that are integral to both the company’s cost structure and its product distribution. The S-1 will be required to disclose how those relationships translate into accounting revenue, and whether revenue recognition reflects genuine third-party demand or, in part, transfers of resources between strategic partners.

The fourth will be the path to profitability. Anthropic is reportedly approaching its first quarter of operating profit, according to Wall Street Journal reporting in May 2026 (Hyatt & Schechner, 2026). The S-1 will reveal whether that profitability is durable or coincidental, and whether the company can sustain it through the inevitable compute cost increases of the next model generation.

The Bottom Line

References