Anthropic told investors this week that it expects to generate $10.9 billion in revenue in the second quarter of 2026, more than double the $4.8 billion it reported in Q1, and to post its first-ever operating profit of $559 million in the process. The Wall Street Journal broke the story on May 20, with CNBC and Bloomberg independently confirming the figures. The numbers were shared as part of an ongoing funding round that could value the company above $900 billion, surpassing OpenAI’s $852 billion valuation from March.
The growth rate is legitimately staggering. Less than a year ago, Anthropic was telling investors it did not expect full-year profitability until at least 2028. Now it is projecting an operating profit in a single quarter at roughly the same point in the calendar. For context, the quarterly revenue trajectory outpaces what Google and Facebook reported in the periods leading up to their own IPOs.
What Is Driving the Revenue
The primary engine is enterprise adoption of Claude Code, Anthropic’s agentic coding assistant. The number of customers spending over $1 million annually on Claude doubled from roughly 500 to more than 1,000 between February and April 2026 alone. Bristol Myers Squibb announced this week that it is deploying Claude across more than 30,000 employees for drug discovery, manufacturing optimization, and regulatory documentation. That deal is representative of the broader shift: enterprises are no longer running chatbot pilots. They are wiring Claude into core workflows.
CEO Dario Amodei acknowledged the pace of growth publicly at a developer conference earlier this month, saying the company had planned for 10x annual growth but actually saw 80x in Q1. That is not a marketing line. It is an operational warning about infrastructure strain.
The Profit That Requires a Footnote
Here is where the Anthropic first profitable quarter Q2 2026 story gets more complicated.
The $559 million operating profit figure includes model training costs but excludes stock-based compensation. More importantly, the Wall Street Journal noted explicitly that it is unclear what accounting methods Anthropic has used, since the company is not yet subject to public company reporting requirements. These are projections shared with investors during a fundraising process, not audited results.
There is also a structural issue embedded in the numbers. According to SpaceX’s S-1 filed with the SEC, Anthropic is paying SpaceX $1.25 billion per month for compute access to the Colossus data center in Memphis through May 2029. That is roughly $15 billion per year in compute costs at full rate. However, the SpaceX S-1 also discloses that Anthropic receives a reduced fee during the May and June ramp-up period as it transitions onto the new infrastructure.
In other words, the two months in which Anthropic is claiming an operating profit happen to be the same two months in which its largest compute bill is discounted. Anthropic’s own guidance acknowledges the company may not remain profitable through the rest of the year once full compute spending resumes.
This does not mean the revenue growth is manufactured. The enterprise adoption driving $10.9 billion in Q2 is real and backed by multiple independent sources. But the profitability milestone is being announced under conditions that are unlikely to repeat in Q3 or Q4.
The IPO Context
The timing of this disclosure is not accidental. Anthropic is actively raising a new round at over $900 billion, and investors were reportedly asked to submit allocations within 48 hours. On the same day the revenue figures leaked, Bloomberg reported that OpenAI is preparing to confidentially file its IPO S-1 with the SEC as early as this week, targeting a September listing.
Both companies are racing to define their valuation narratives before the other reaches public markets. Whoever goes public first sets the comparable for the other. A projected $10.9 billion quarter, even with footnotes, is a much stronger fundraising document than last year’s guidance.
What This Actually Tells You About Enterprise AI
Strip away the fundraising context and there are two things worth taking seriously.
First, enterprise AI adoption has crossed a threshold. Customers are not experimenting. They are committing at the $1 million-plus annual level in numbers that were unimaginable 18 months ago. Bristol Myers is not a startup running a Claude pilot. It is a $120 billion pharmaceutical company deploying AI into drug submissions and manufacturing decisions.
Second, the compute cost structure remains the central unresolved problem in frontier AI. Anthropic paying $1.25 billion per month to SpaceX is not a sustainable long-term unit economics story unless revenue continues compounding at the current pace. Dario Amodei’s “80x growth” comment is impressive, but it also explains why Anthropic needed to buy access to Elon Musk’s data center in the first place.
The Anthropic $10.9 billion revenue Q2 projection is real. The Anthropic operating profit $559 million is real but conditional. The compute pressure will return in Q3. Whether the revenue trajectory continues to outrun it is the only question that matters for the rest of 2026.
