Media reports citing informed insiders indicate that Anthropic PBC, a global leader in AI models and applications, expects its initial public offering (IPO) on the U.S. stock market to match or potentially exceed the record-setting scale of SpaceX's June listing. This latest development in the AI supply chain underscores the intense demand from both institutional and retail investors hoping to capitalize on the unprecedented wave of artificial intelligence investment. According to sources, the developer of the Claude AI application platform is currently conducting valuation and fundraising assessments, with management preparing to publicly submit necessary documents for a potentially massive IPO as early as the end of this month. They also noted that Anthropic's Chief Financial Officer, Krishna Rao, deliberately avoided valuation discussions during a recent investor presentation.
Notably, in May of this year, Anthropic completed a funding round valued at approximately $65 billion, reaching a valuation near $1 trillion (around $965 billion), surpassing competitor OpenAI's $852 billion valuation from its March fundraising. Data compiled by institutions shows that Elon Musk's rocket and satellite company, SpaceX, initially raised $75 billion, making it the largest initial stock offering in history, with a market capitalization of $1.77 trillion upon listing. Including the so-called "over-allotment option," the final fundraising amount increased to an unprecedented $86.2 billion, an option typically exercised if the stock rises in early trading. Insiders noted that discussions are ongoing, and details such as the scale of Anthropic's IPO could still change, with the sources requesting anonymity as the information has not been made public. A representative for Anthropic did not immediately respond to requests for comment.
Claude's Near-Trillion-Dollar Ascent to Capital's Summit
Anthropic's more ambitious IPO target compared to SpaceX reflects how leading companies in the AI industry are reshaping the technology investment landscape. This company, founded just five years ago, raised $65 billion at a $965 billion valuation in May, exceeding OpenAI's $852 billion valuation achieved during its $122 billion funding round in March. Although Anthropic's adjusted operating profit turned positive faster than market expectations in the second quarter, the company recorded a net loss of nearly $42 billion in 2025, roughly five times the prior year's net loss of about $8.3 billion, according to a document reviewed by institutions. On the positive side, demand for AI applications is rapidly increasing, as evidenced by the sharp rise in Anthropic's revenue figures. Reports citing informed insiders indicate that preliminary second-quarter revenue exceeded $11.5 billion, compared to just $787 million in the same period of 2025, with annualized revenue run rate reaching a staggering $65 billion by the end of July. The revenue run rate is a metric that projects annual revenue based on a shorter period's performance.
Anthropic's underlying business is no longer a single chatbot but a full-stack commercial system encompassing "frontier models + API + Agent Harness + enterprise workflows + multi-cloud distribution": Claude provides core intelligence, Claude Code targets software development as a high-frequency, quantifiable return scenario, Model Context Protocol (MCP) and Agent SDK connect enterprise data and tools, and the AI agent super-platform Claude Cowork extends the same agent capabilities to documentation, compliance, customer service, and knowledge work, reaching enterprise-scale procurement channels through Amazon Bedrock, Google Vertex AI, Microsoft Azure, and direct APIs. The AI agent platform Claude Cowork, closely focused on AI programming applications, is generating strong compounding returns first because code compilation, test passing, and tool call success can all be automatically verified. Anthropic also converts real failure trajectories into evaluation sets (Evals) and next-generation model training objectives, forming a collaborative design loop of "user usage—product feedback—model enhancement—experience improvement—usage growth."
Demand for large models has shifted from one-time Q&A to multi-step workflows where AI agents continuously handle planning, retrieval, code execution, tool calls, and result verification: each task exponentially consumes tokens while accumulating real usage data and evaluation feedback, driving further improvements in model capability, task success rates, customer stickiness, and revenue, creating a compounding loop of "more calls—stronger models—higher automation rates—more paid calls." Claude Cowork and its plugins for legal, sales, marketing, and data analysis are the concentrated embodiment of this paradigm—it elevates Claude from a chat interface to the "execution layer" of enterprise work, triggering market concerns about contraction of traditional software seats, commoditization of features, and weakened pricing power. The related release triggered an evaporation of approximately $285 billion in market value across the software and data services sector, with Goldman Sachs' U.S. software stock basket falling 6% in a single day; while this impact does not mean traditional SaaS will be immediately replaced, it marks a new phase where global AI competition has shifted from "who has the strongest model" to "who can make agents take over the most workflows and convert tokens into cash flow."
Some insiders indicated that Anthropic and its competitors are grappling with the extremely high costs of building more advanced AI super-applications. Training so-called frontier models requires massive AI computing infrastructure resources—in just one agreement with data center operators, Anthropic agreed to deploy AI computing resources on a massive scale through SpaceX, with contract values potentially reaching tens of billions of dollars over the next three years. According to media reports, Anthropic is expected to reach the public market before OpenAI, which is currently considering a 2027 listing. Both companies have separately filed confidential IPO applications. Insiders previously revealed that before publicly filing offering documents, Anthropic is finalizing a revolving credit facility that will exceed the initial target of approximately $10 billion. Insiders stated that Anthropic is working with Morgan Stanley, Goldman Sachs Group, and JPMorgan on the IPO, with other major commercial banks including Citi potentially joining the underwriting team. One insider said Anthropic is considering adopting super-voting shares to give CEO Dario Amodei, who holds about 2% of the company, and other co-founders stronger effective control. The Information first reported on the super-voting share plan.
Data compiled by institutions shows that if Anthropic's total IPO scale exceeds SpaceX's, it would easily make 2026 the best year on record for U.S. IPO fundraising. As of August 19, newly listed U.S. companies this year have collectively raised $160.6 billion, below the record $195.2 billion set in 2021. Anthropic's listing will further bolster what has already been a stellar year for global equity IPO markets. This year has already produced two of the largest public offerings in history, including SpaceX, the U.S. super-tech giant founded by Musk focused on "AI + space exploration," which made a blockbuster debut with a record IPO, and South Korea's memory chip giant SK Hynix, which raised $26.5 billion through its initial offering of American Depositary Receipts.
Seven-Month Revenue Surge of 7.2x: Anthropic Challenges SpaceX's Fundraising Throne with Agent Compounding
Anthropic's financial trajectory provides rare verifiability for Wall Street's extremely optimistic earnings path and record IPO expectations. Anthropic's annualized revenue run rate climbed from approximately $9 billion at the end of 2025 to $47 billion in May 2026 and over $65 billion by the end of July, representing a revenue scale of approximately 7.2 times within seven months, an increase of about 622%. Preliminary second-quarter revenue exceeded $11.5 billion, approximately 14.6 times the $787 million in the same period of 2025, and also about 2.4 times the first quarter's $4.73 billion, achieving positive adjusted operating profit for the first time. These fundamental operating metrics indicate that AI inference-related revenue and enterprise seat revenue are beginning to grow faster than R&D, personnel, and unit computing costs in the AI inference era. Model compression, caching, batching, routing across different models, and custom chip deployment continue to reduce per-token costs, while high-paying enterprise workloads increase revenue per unit of computing. However, the $65 billion is an annualized run rate extrapolated from end-of-July revenue velocity, not realized full-year revenue, and different companies may use different gross or net recognition standards for cloud channel revenue.
Anthropic's potential to match or exceed SpaceX's $75 billion initial IPO, and even the $86.2 billion final scale including over-allotments, is essentially the result of "earnings growth certainty, scarcity premium in the AI industry, and strong capital demand" working together: the company just raised $65 billion at a $965 billion valuation in May, more than doubling its valuation from $380 billion in February, and offering the market an extremely scarce pure-play frontier AI investment opportunity before OpenAI. Wall Street giants like Morgan Stanley, Goldman Sachs, and JPMorgan can leverage structural AI allocation demand from sovereign funds, pension funds, and large tech funds to absorb the massive issuance. Meanwhile, the IPO is not just a market reward for Claude but also capital raising for training clusters, long-term inference capacity, talent, and potential acquisitions: its pre-listing revolving credit facility is expected to exceed $10 billion, further indicating the company is building a "equity + credit + long-term computing contracts" financing stack in advance. Based on a rough calculation at the $965 billion valuation, $75 billion to $86.2 billion represents approximately 7.8% to 8.9% of the current valuation; the actual float ratio and equity dilution will depend on the IPO valuation and the mix of new and existing shares.
Anthropic possesses the clearest path to commercialization and profitability for frontier models to date, but it cannot simply be valued like a mature software company after listing: if Claude Code and Cowork continue to expand and unit token costs keep declining, the massive IPO could become the capital starting point for the next round of enterprise AI platform expansion; if model price wars, computing costs, customer multi-model procurement, or revenue recognition standards weaken growth quality, valuations based on 2028 forward revenue could also compress sharply. Earlier media reports quoted two individuals familiar with Anthropic's financials as saying the company's annual revenue is expected to reach $190 billion to $200 billion by 2028. Therefore, Wall Street is truly betting not on Anthropic's current profits but on whether it can extend the "model—product—feedback—training" compounding across the entire enterprise AI agent market. Based on the market's currently discussed $2 trillion valuation and $200 billion revenue outlook, this corresponds to approximately 10 times 2028 sales; in comparison, the $965 billion valuation corresponds to approximately 14.8 times the current run rate, indicating that growth realization could actually digest some of the valuation multiple.
Comments