Anthropic, the AI giant behind the Claude model, has seen its revenue surge roughly twelvefold year over year to nearly $4.6 billion in 2025, according to IPO filings viewed by internal sources.

The company's public offering could push its overall valuation beyond $2 trillion, surpassing the approximately $1.77 trillion valuation achieved by SpaceX at its listing.

With a planned fundraising target approaching $100 billion, Anthropic is positioned to exceed SpaceX's roughly $85.7 billion in proceeds after over-allotment, potentially making it the largest initial public offering in history.

Where the growth is coming from

As cutting-edge AI agents such as Muse and Astra rapidly gain global traction, AI is accelerating its shift from conversational tools toward automated coding, research, and the execution of complex enterprise workflows, simultaneously expanding application revenue and AI compute investment.

Anthropic's IPO aims to bring this expansion into the global public capital markets. Sources indicate that beyond the roughly twelvefold revenue jump, the IPO prospectus also reveals that Anthropic will assume obligations totaling up to $518 billion over the coming years for cloud services, AI computing capacity, and broad AI infrastructure spending.

A potential listing valuation exceeding $2 trillion means investors will be directly pricing the commercialization capability of frontier models, compute procurement commitments, and long-term profitability prospects, with implications extending to OpenAI and related AI compute infrastructure suppliers.

Recent advances in frontier AI agents and large models represented by Muse, Astra, and Anthropic's own technologies have provided an important technical foundation for the large-scale commercial expansion of AI applications across industries and the continued surge in AI compute demand.

In particular, the broadening scope of agent applications is expected to simultaneously boost demand for AI super-accelerators such as AI GPUs and TPUs, high-performance CPUs, high-performance HBM, DRAM, and NAND memory chips for data centers, and high-speed optical interconnect components.

The scale of the offering and its financial picture

Anthropic is seeking to raise up to $100 billion at a valuation of approximately $2 trillion, though the final plan may still be adjusted in conjunction with major Wall Street banks, and whether it sets a fundraising record will depend on the final offering size.

The IPO prospectus also shows that of the company's approximately $42 billion net loss in 2025, about $34 billion stems from accounting charges related to the appreciation of financing instruments that may convert into shares.

Operating losses for the year exceeded $8 billion, with compute and infrastructure spending reaching $7.33 billion. The $518 billion figure represents multi-year future spending obligations and cannot be directly treated as current cash burn or on-balance-sheet interest-bearing debt.

Meanwhile, Anthropic disclosed in May that its annualized revenue run rate had surpassed $47 billion, indicating continued commercial expansion, though this metric cannot be equated with revenue actually realized for the full year.

Public markets put AI commercialization to the test

Anthropic's prospectus presents a grand AI vision, but costs are climbing sharply. According to the IPO filings, Anthropic is making an unprecedented technology bet: that artificial intelligence will transform the global economy more profoundly than industrialization, electricity, and the internet combined. But the cost of achieving this will be enormous.

The prospectus shows Anthropic recorded a $42 billion net loss in 2025 and plans to spend $518 billion over the coming years to meet obligations in cloud services, computing, and infrastructure.

According to documents first disclosed by Reuters, Anthropic's revenue in 2025 grew twelvefold from the prior year to nearly $4.6 billion. At the same time, after excluding the impact of various liability write-downs mainly related to prior financing, the company's operating-level loss still exceeded $8 billion.

The public offering could value the company at more than $2 trillion, leveraging the rapid rise of this five-year-old AI startup lab and making it a benchmark for Wall Street to assess the value of leading AI companies, including rival OpenAI.

As it advances its listing plans, Anthropic is confronting evidence from its own research: increasingly autonomous AI models may exhibit unexpected and potentially harmful behaviors, including sabotaging code in controlled tests, assisting fraud, and manipulating information.

These findings have entered broader public discussion, raising concerns about how companies can ensure these increasingly powerful systems remain under control as they race to deploy them commercially.

Anthropic CEO Dario Amodei has called on the global AI community to slow the pace of releasing new capabilities in response to these concerns. Nevertheless, ahead of its own expected IPO, Anthropic last week launched its new Opus 5.5 model to counter momentum generated by OpenAI since the release of GPT-6 Astra.

The AI lab spent $7.33 billion on compute and infrastructure last year, triple the 2024 figure and more than half of its $12.65 billion in total operating expenses.

Sources previously indicated that Anthropic's formal U.S. IPO timeline could be pushed back to after the November U.S. midterm elections. The offering will give public investors a stake in the AI race, which until now has been funded primarily by venture capital firms, sovereign wealth funds, and large technology companies.

Anthropic declined to comment on any IPO news. The IPO prospectus also shows the company expects a valuation of a stunning $2 trillion. Anthropic's targeted valuation is more than double its own May estimate of $965 billion.

As the chart above illustrates, while AI application-related revenue has surged, Anthropic's 2025 losses have also surged. As the valuation soared, increasing the value of shares promised to investors, Anthropic's books recorded massive losses in tandem.

Of the nearly $42 billion net loss, about $34 billion consists of accounting charges. This reflects the estimated value increase of financing instruments that may eventually convert into Anthropic shares, not cash spent on day-to-day operations.

Anthropic said nearly a quarter of last year's revenue came from two customers. The company also warned in its risk factors that many of its largest customers are not bound by long-term contracts and may cut or stop spending.

The prospectus shows that as of December 31, Anthropic held $20.28 billion in cash, cash equivalents, and short-term investments.

Anthropic will list following SpaceX's recent large IPO. That offering valued Elon Musk's company at $1.77 trillion. SpaceX shares rose 19% to $160 on their June 12 debut and currently trade around $147, still above the $135 IPO price.

This performance may make investors more cautious when assessing still-elevated valuations of high-growth companies. AI and chip stocks have recently suffered a sell-off, and given Anthropic's ambitious growth projections, its offering will further test whether market enthusiasm for AI trades can withstand stricter scrutiny.

Despite high interest rates, economic uncertainty, and market concerns about valuations, the listing is still expected to cap one of the strongest years for the U.S. IPO market since 2021.

As the chart above shows, Anthropic's revenue has surged as frontier AI applications and AI agents accelerate adoption among enterprises worldwide, with revenue growing more than tenfold from 2024 to 2025.

Anthropic's biggest competitor remains OpenAI, and the two companies are engaged in fierce competition for enterprise customers, talent, and influence in Washington. OpenAI confidentially filed for an IPO in June. According to media reports, the ChatGPT developer is expected to go public by early 2027 or sooner.

Anthropic was founded by a group of researchers who left OpenAI after disagreements over corporate governance and AI safety issues. Anthropic released its first large language model in March 2023, entering direct competition with OpenAI.

It also competes fiercely with SpaceX's xAI, Alphabet's Google, and Facebook parent Meta in accelerating the buildout of AI compute infrastructure critical to the accelerated penetration of AI applications and AI agents.

Analysts expect the first company to list will establish a valuation benchmark for the AI industry and attract investors who have waited years to bet on the technology through direct investments in AI-focused companies.

Amazon and Google were two important early strategic partners of Anthropic. Both companies invested billions of dollars in the startup while providing the cloud infrastructure needed to train and deploy Claude models.

Anthropic and its CEO Amodei have clashed with the White House over the use of its tools, leading the Pentagon to briefly blacklist Anthropic; a U.S. judge blocked the measure in August.

AI kicks off a capital relay race

Muse, the released GPT-6 Astra, and Claude's new-generation models are driving a shift toward handing more actual work to AI. From a system architecture perspective, a task may involve planning, retrieval, code execution, tool calls, and verification: GPUs or specialized accelerators handle model computation, CPUs handle browsers, sandboxes, and scheduling, HBM and system memory support models and runtime states, and storage preserves files, databases, and long-term memory.

Meta's latest technical materials explicitly state that Muse uses independent cloud virtual machines configured with CPU, memory, storage, and browsers, and supports parallel sub-agents. These underlying AI technology trends all point to infrastructure demand for agent proliferation, spanning both model inference and cloud execution environments.

The chart above shows the scale of AI computing resource spending officially provided by Anthropic, which clearly demonstrates a surge. Storage chip components in AI data center server clusters remain the clearest supply bottleneck in the AI compute supply chain.

Market research firm TrendForce estimates that server DRAM contract prices will cumulatively rise about 270% in 2026, with enterprise SSD prices rising about 235%. In 2027, HBM contract prices may still rise 70% to 140%, continuing to show doubling-style growth. These data reflect the combined effect of sustained AI compute demand expansion and memory chip price increases.

TrendForce's latest estimate shows that DRAM and NAND together will account for 47% of major cloud service providers' capital expenditure in 2026, rising to 68% in 2027, driven by both increased procurement volumes and higher prices.

Application expansion also benefits from declining costs as AI agents efficiently complete complex tasks and conduct deep scientific research. Anthropic said Opus 5.5, launched on September 22, reduces costs by about 40% on typical workloads compared with Opus 5. Sonnet 5.5, released on September 28, further covers everyday tasks, with per-task costs reduced by up to about 30% in testing.

This efficiency improvement is expected to make more work commercially viable, while the increment in total compute demand depends on whether new users and task volumes can outpace the decline in resource consumption per task.

The previous rally in the AI "neocloud" investment sector, led jointly by CoreWeave and Nebius, also had support from demand and pricing. CoreWeave disclosed on September 17 that it continued to sign new compute contracts at higher prices in the third quarter, including short-term contracts with terms of about three to six months.

This record potential IPO has been described by some Wall Street analysts as a "capital relay" for the global AI supply chain: public market funds pick up where early investments left off, providing support for the time gap between locking in compute capacity in advance and future customer payments.

However, some analysts emphasize that Anthropic's customer spending plans may be flexible, while infrastructure commitments span multiple years, so investors need to assess whether customer retention, inference service gross margins, procurement fulfillment pace, and cash recovery can match up.

If the IPO is completed successfully, it will help significantly strengthen Anthropic's compute procurement capability and boost confidence in demand realization across the core AI compute supply chain. What ultimately supports the valuation is Claude's ability to continuously convert technical capability into customer payments and operating cash flow.