Anthropic posts $16.8 billion Q3 revenue, claims adjusted profitability for first time
Artificial intelligence company Anthropic delivered a striking report card to potential investors at a pivotal moment in its IPO preparations: third-quarter 2026 revenue reached $16.8 billion, up approximately 45% quarter-over-quarter, with the company claiming profitability on an adjusted basis and positive quarterly cash flow. This marks the first time the company has disclosed relatively complete financial data to public-market investors, with the disclosure made during an investor day event held Wednesday local time at its San Francisco headquarters.
Anthropic's revenue trajectory shows a steep growth curve. First-quarter revenue this year was approximately $4.7 billion, climbing to $11.6 billion in the second quarter, and further rising to $16.8 billion in the third quarter. In just two quarters, quarterly revenue expanded roughly 2.6-fold, several times higher than the approximately $4.6 billion in revenue for all of last year. Simply annualizing the third-quarter figure yields a corresponding revenue run-rate of approximately $67.2 billion.
However, there is a critical qualifier behind these impressive numbers: "adjusted." The profitability Anthropic claims is on an adjusted basis, not net income calculated under U.S. Generally Accepted Accounting Principles (GAAP). The company has not yet disclosed specific adjustment items, leaving outsiders unable to determine which one-time costs or non-cash items were excluded.
According to the Financial Times, Anthropic emphasized rapid growth, efficiency improvements, and a safety-first development strategy to attendees at the investor day, hoping to garner support for its planned IPO. The report also noted that chips and data centers remain the company's largest expense items, while the company told investors that third-quarter cash flow had turned positive.
Bloomberg, citing sources familiar with the matter, reported that Anthropic's third-quarter adjusted gross margin reached 85%, adjusted operating margin exceeded 11%, and free cash flow was also positive. Based on $16.8 billion in revenue and an 11% operating margin, adjusted operating profit would be approximately $1.85 billion.
Scrutiny of profit metrics and the compute ledger
The claim of adjusted profitability immediately sparked broader discussion about the financial sustainability of AI companies. Anthropic has not yet disclosed detailed profit figures, while previously leaked documents showed the company posted a net loss of nearly $42 billion in 2025, with revenue of approximately $4.6 billion and an operating loss exceeding $8 billion that year. The transition from massive losses to claimed adjusted profitability in just a few quarters is a turnaround whose speed itself requires more complete disclosure to substantiate.
The cost structure of the AI industry has its own peculiarities. Revenue growth for model providers is typically accompanied by simultaneous expansion in compute, chips, data centers, and cloud services. Anthropic has already signed extensive long-term cloud computing and infrastructure agreements, with future spending commitments reaching hundreds of billions of dollars. The Financial Times reported that the company plans to invest more than $500 billion in infrastructure and compute over the next decade. These long-term contracts mean that even if a particular profit metric turns positive in a given quarter, capital expenditure pressure over the coming years remains enormous.
One industry insider noted that evaluating an AI company's operating condition requires examining at least four separate ledgers: the training ledger, the inference ledger, the channel ledger, and the cash ledger. Training costs are concentrated in the model development phase, inference costs grow linearly with customer usage volume, channel revenue-sharing affects gross margin calculations, and the cash ledger involves depreciation, prepayments, and the mismatch between long-term contract commitments and actual payment timing. A single quarter of adjusted profitability is insufficient to conclude that the business model has been fully validated.
IPO valuation and market expectations
Anthropic is accelerating its IPO timeline. Bloomberg previously reported that the company could formally launch its IPO roadshow as early as the week of November 9, aiming to begin trading before Thanksgiving, though the specific timing remains subject to change. Reuters reported that the listing is expected to occur after the U.S. midterm elections in November.
On valuation, some investors believed prior to this week's investor day that a valuation of approximately $1.8 trillion to $2 trillion for Anthropic was reasonable. The Wall Street Journal previously reported that the company's IPO could raise up to $100 billion at a valuation of approximately $2 trillion. If it approaches that scale, it would surpass the $75 billion raised by SpaceX in June 2025, becoming one of the largest initial public offerings in global history.
But market expectations are also rising. The Wall Street Journal, citing investor expectations, reported that Anthropic may need to reach an annualized revenue run-rate of over $110 billion by year-end to satisfy some of the market expectations already formed ahead of the listing. This means the company needs to maintain an extremely high quarter-over-quarter growth rate in the fourth quarter—a requirement that becomes increasingly difficult as the base continues to expand.
Competitive landscape and safety concerns
Anthropic was founded by former OpenAI executives, and its Claude series of models competes directly with OpenAI's ChatGPT. Alphabet's Google and Amazon are both important investors and compute partners, while also competing with Anthropic in the broader AI market.
Competitive pressure is intensifying. Meta recently released new AI models, claiming performance in programming, finance, and legal tasks comparable to ChatGPT and Claude. As multiple players continue investing in model development, whether Anthropic can sustain its current growth rate is one of the key variables determining IPO success.
AI safety issues also constitute a potential liability. In its draft IPO filing documents, Anthropic devoted roughly one-third of the content to risk factor disclosures, including the possibility that AI systems could exhibit unpredictable behavior or even pose serious risks to humans. The company has identified certain anomalous behaviors of AI agents in test scenarios within research environments, but stated it has not observed evidence of such deviations in actual deployments.
For the broader AI industry, Anthropic's third-quarter data carries indicative significance. It demonstrates that generative AI has acquired the capacity to generate large-scale commercial revenue in the short term, but whether that revenue can be reliably converted into net profit and free cash flow remains the core question that public markets will need to continuously validate. Anthropic is expected to publicly release its IPO documents in the coming weeks, at which point the calculation methodology for adjusted profit, cash flow structure, and details of long-term compute commitments will be revealed one by one.
In an environment where interest rates remain elevated and financing costs for technology companies are rising, the valuation logic for AI companies is shifting from a pure revenue-growth narrative toward a more granular examination of unit economics and return on capital. Anthropic's listing process will serve as an important stress test of public-market appetite for AI infrastructure investment.