Five years after former OpenAI researchers founded the venture over safety concerns, Anthropic priced its shares at $83 on Wednesday and opened at $111.50 on Thursday, briefly valuing the company at around $2 trillion.
The public market debut on the Nasdaq under the ticker ANT replaces private funding rounds with daily mark-to-market pricing. It sets up an immediate operational question for technology leaders: can a software firm running on $17 billion in quarterly revenue sustain $900 billion in committed compute capacity if enterprise software budgets contract?
The Scale of the Listing
Anthropic raised a record $81.3 billion in the initial public offering, making it the first of the leading model developers to list on Wall Street. That initial sum surpasses the $75 billion that SpaceX secured during its June IPO at a valuation of $1.77 trillion. SpaceX later expanded its haul to $86 billion when underwriters exercised an option to buy additional stock. Anthropic granted a similar overallotment option to its underwriters, which would push total proceeds to about $93.5 billion.
Anthropic shares opened 34% higher on Thursday before settling minutes into trading at around $98, which left the business valued at about $1.8 trillion. At the IPO price of $83, which had already exceeded the expected range, the creator of the Claude chatbot stood at roughly $1.5 trillion when counting employee stock options and restricted shares. Chief executive Dario Amodei addressed the market directly from the trading floor.
"Today puts the company in a stronger financial position, and in doing so, only strengthens our resolve to do the right thing,"
Amodei made that statement at the Nasdaq in New York's Times Square, pointing to safe AI as a primary corporate objective. Daniela Amodei, the company's president who founded Anthropic in 2021 with Dario Amodei and several colleagues, told CNBC that the listing represented "one stop on the journey" and noted that the firm would remain the same operational entity the following day.
Revenue Growth Against Compute Liabilities
The business baseline behind the valuation rests on top-line expansion. Anthropic generated $17 billion in revenue in the third quarter, compared to $1.4 billion a year earlier. Much of that cash intake comes from enterprise customers paying to deploy Claude models for computer coding workloads.
Yet the filings reveal the underlying financial architecture of large language model operations. Anthropic disclosed roughly $900 billion in commitments for computing capacity. The regulatory prospectus warned that the company must make many of these scheduled payments even if end-user demand falls short of internal targets. This creates a severe structural imbalance: with $17 billion in quarterly revenue (translating to an annualized run rate of $68 billion), the firm's incoming cash flow covers only a fraction of the monumental $900 billion infrastructure liabilities under contract.
Directly analyzing whether enterprise software budgets can sustain this load under contract constraints reveals a fragile equilibrium. Enterprise technology spending is finite, and traditional software budgets are already being cannibalized to fund AI deployments. If macroeconomic pressures or slower-than-expected productivity gains force CIOs to trim software outlays, enterprise software budgets will simply fail to absorb the pricing required to service $900 billion in fixed hardware and cloud commitments. Because Anthropic's compute contracts carry mandatory take-or-pay or minimum volume thresholds, a downturn in downstream developer adoption would leave the company legally bound to billions in hardware payments without the corresponding top-line cash generation, transforming fixed infrastructure assets into a severe liquidity hazard.
"This may be the breath of fresh air that the tech trade needed to move beyond the fog of artificial intelligence that we've been living in for the last couple of weeks,"
Art Hogan, chief market strategist at B. Riley Wealth Management, described the market appetite with that assessment in comments to AFP. The listing proceeded through volatile market conditions for tech shares and corporate disruptions at suppliers. Chip partner Nvidia, an investor in Anthropic and dozens of other AI companies, faced turmoil after chief executive Jensen Huang resigned following a reported relationship with an alleged intelligence asset.
Structural Risks and Safety Mandates
Public market pricing answers whether near-term revenue covers the $900 billion computational load: it does not yet do so from operational cash flows alone, shifting the burden onto ongoing market liquidity and uninterrupted corporate uptake of coding models. Anthropic outlined systemic exposures in its IPO filing, warning that advanced systems "could pose catastrophic or existential risks to society if not developed and deployed responsibly." Outside the exchange, about a dozen protesters held signs, and academic critics issued a public statement titled Misanthropic, arguing the business uses safety branding to court institutional capital while pursuing aggressive deployment.
Audit third-party vendor agreements this week to identify whether multi-year commitments to AI coding infrastructure include mandatory minimum volume thresholds if downstream enterprise software usage drops.
How the picture shifted
Five years ago, former OpenAI researchers set out to build a safety-focused startup on venture checks, but now Anthropic is storming the Nasdaq with a valuation near two trillion dollars. The developers' operational appetite grew just as fiercely: by early November, internal usage of Claude expanded over 250-fold, turning compute infrastructure into the ultimate cash drain. Wall Street is effectively betting that enterprise programmers writing code with Claude will generate cash faster than data center bills accumulate. Can Dario Amodei keep up with mandatory infrastructure commitments if corporate clients decide to trim their software budgets?
Our stories: Claude Inference Costs Surpass Anthropic Payroll (November 2).