Anthropic Plans to Advance First IPO in November, Target Valuation of Approximately $2 Trillion
According to reports, AI company Anthropic plans to advance its initial public offering (IPO) in November, targeting a valuation of approximately $2 trillion, with a maximum fundraising scale of up to $100 billion through this IPO. The originally scheduled October listing window has been postponed to November to allow the company to disclose its third-quarter financial performance, further substantiating its market competitive position.
Anthropic's valuation has seen a steep increase in less than a year: in February 2026, the company completed a $30 billion financing round with a valuation of $380 billion; by May 2026, it completed another $65 billion financing round, raising its valuation to $965 billion; if this IPO ultimately achieves a $2 trillion valuation, it would mean the company's valuation has grown by about five times in less than a year.
Supporting this valuation leap is the rapid expansion of revenue: Anthropic's annualized revenue has grown from approximately $9 billion at the end of 2025 to about $65 billion by the end of July 2026; according to estimates from asset management company ARK Invest in August this year, the company's current annualized revenue run rate has exceeded $74 billion (this data is based on third-party information and company disclosures, not audited financial data); the company internally expects annualized revenue to reach about $120 billion by the end of 2026, further rising to about $200 billion by 2028.
At a $2 trillion valuation, this corresponds to about 31 times the current annualized revenue, about 17 times the expected revenue by the end of 2026, and about 10 times the expected revenue by 2028—this valuation multiple is considered significantly higher than traditional software industry valuation levels, especially since the company has not yet disclosed audited financial statements.
Reports indicate that NVIDIA is considering a cornerstone investment of $10 billion in this IPO; Amazon and Google have previously deeply bound themselves to Anthropic through dual roles of investment and computing power supply—Amazon has committed to investing over $100 billion in AWS computing resources for Anthropic's cloud infrastructure over the next decade, while Google has committed to selling multiple gigawatt-scale TPU chip computing power to it. Analysts believe this "both shareholder and supplier" structure means these strategic investors have additional profit channels and information advantages that ordinary public investors do not possess.
In terms of governance structure, Anthropic has established a "Long-Term Benefit Trust" mechanism, which will ultimately control the majority of the company's board seats, designed to prioritize the principles of AI safety rather than merely pursuing shareholder returns, meaning that public shareholders' voting rights will be substantially limited after the IPO. Additionally, the company's currently disclosed "adjusted operating profit" metric does not account for revenue sharing paid to distribution partners or ongoing model training costs—since cutting-edge large models require continuous retraining, this fundamentally differs from the traditional software business model of one-time development and long-term low-cost maintenance, raising questions about whether the disclosed gross margin of over 80% accurately reflects its profitability. Meanwhile, CEO Dario Amodei has long advocated for "slowing down the pace of AI development" to address potential risks, which has also raised concerns among some investors that the company's future financial outlook may be affected by changes in the pace of technological development.
In comparison, competitor OpenAI is currently expected to initiate its IPO in 2027, with a previous financing round negotiating a valuation expectation of over $1.2 trillion, having recently completed over $120 billion in financing; reports indicate that perpetual contracts around Anthropic's equity have been trading on some cryptocurrency exchanges since June this year, reflecting the market's heightened attention to this potential IPO extending beyond the traditional stock market.
Source: Public Information
ABAB AI Insight
Anthropic's financing pace over the past year has exhibited a typical "AI arms race" acceleration characteristic—from the $30 billion financing in February 2026 corresponding to a $380 billion valuation, to the $65 billion financing in May corresponding to a $965 billion valuation, and now aiming for a $2 trillion IPO valuation, the company has completed unprecedented financing rounds almost every two to three months. This density of financing and speed of valuation increase have historically only occurred during the internet bubble and in the recent generative AI wave, reflecting that capital competition for leading AI labs has reached an almost cost-ignoring fever pitch.
It is noteworthy that the capital flow surrounding Anthropic is not merely financial investment but is deeply embedded in the computing power supply chain—Amazon has committed to investing over $100 billion in AWS computing over ten years, Google has committed to selling multi-gigawatt TPU capacity, and NVIDIA is considering a $10 billion cornerstone investment in the IPO, with these three tech and chip giants simultaneously binding themselves to Anthropic in a "investor + supplier" dual role. The essence of this funding path is that cloud computing and chip giants are transforming capital expenditures originally intended to expand their own computing capacity into equity investments in leading AI labs, locking in future computing procurement orders while sharing potential valuation growth benefits, creating a structure of capital and computing power that mutually reinforces and cycles.
A comparable historical case is the "vendor financing" model common between telecom equipment vendors and emerging operators during the late 1990s telecom infrastructure bubble—equipment vendors helped customers purchase their equipment through investment or credit, boosting both parties' short-term performance but also embedding chain risks when cycles reverse. The current AI industry is in a phase where leading labs are racing toward IPOs, with deep binding of capital and computing power, and the valuation competition between Anthropic and OpenAI (the latter also expected to reach a valuation of over $1.2 trillion at IPO) further illustrates that the entire industry's financing narrative has expanded from "technology capability competition" to a competition of "capital scale and supply chain locking ability."
This event essentially reflects that global capital is highly concentrated toward a very small number of leading AI labs: a target valuation of $2 trillion corresponds to about 31 times the current annualized revenue, indicating that the market is willing to pay a very high premium for "certainty of future growth," and this premium is largely based on the assumption of continued capital and computing power injection from giants like Amazon, Google, and NVIDIA. Mechanically, this concentration of capital can persist because computing power and capital have become scarcer resources in the AI race than technology itself—only a very few companies can simultaneously secure commitments from top cloud providers for computing power and guarantees from top chip manufacturers for capacity, and this scarcity further reinforces the drive for capital to concentrate on these leading companies, forming a typical Matthew effect: the more a company can lock in computing power and capital, the easier it is to obtain more computing power and capital.
ABAB News · Cognitive Law
- Valuation rises faster than revenue, reflecting faith in certainty.
- Both shareholder and supplier, the rules always favor insiders.
- Scarcity has never been technology, but the computing power and capital that can lock in technology.