Anthropic Is Not Going to Hit $100 Billion This Year
Anthropic is not going to hit $100 billion in annualized revenue this year.
Nobody who works in the numbers will say so on the record. The $100 billion figure circulating from The Information is a market extrapolation, not a company projection. Anthropic's own internal forecast for 2026 is $18 billion in revenue, quadruple what it earned in 2025, according to Reuters. The 2027 forecast is $55 billion. Those are the numbers with a source.
The $100 billion figure is what you get when you multiply the current $30 billion annualized revenue run rate by a growth factor derived from the past three consecutive years of 10x annual expansion and then assume the pattern holds through December. It is arithmetically coherent and narratively convenient, the kind of number that sounds rigorous because it involves multiplication.
The more interesting question is not whether Anthropic can reach $100 billion. It is whether the company that was supposed to prove safety and scale could coexist will survive its own growth intact.
Anthropic closed a $30 billion Series G funding round on February 12, 2026, at a $380 billion post-money valuation, per Sacra. That is 27 times annualized revenue for a company that did not exist four years ago. The valuation encodes a bet that Dario Amodei's vision of "a country of geniuses in a datacenter" arrives on a specific schedule and that infrastructure commitments totaling tens of billions of dollars will not be stranded if that schedule slips.
The revenue under the valuation is real. Anthropic grew from $1 billion in annualized revenue in December 2024 to $4 billion by July 2025, $9 billion by December 2025, and $14 billion by the second week of February 2026, per Shanaka Anslem Perera's analysis. It crossed $30 billion in April 2026, according to Tradingkey. Claude Code, its agentic coding product, reached $2.5 billion in annualized billings in approximately nine months of existence, with business subscriptions quadrupling in the six weeks after January 1. More than 500 customers now spend over $1 million annually with Anthropic, up from roughly a dozen two years ago. Eight of the ten largest companies on the Fortune 10 list are Claude customers.
But there are structural details in those numbers that deserve attention.
The first is how Anthropic reports revenue. Anthropic sells through AWS, Google Cloud, and Microsoft Azure. It books the full end-customer spend as revenue and records the cloud partners' cuts as expenses. This is gross-basis reporting, which makes Anthropic's top-line figures larger than they would be under net-revenue accounting, which is how most enterprise software companies report, according to Sacra. Peers who sell through the same cloud channels typically report net. Readers comparing Anthropic's revenue multiples to comparable companies should account for this.
The second detail is the governance structure. Anthropic is a public benefit corporation, or PBC. Its charter legally requires the board to balance shareholder returns against the company's stated mission: responsibly developing and maintaining advanced AI for the long-term benefit of humanity, per the Harvard Law Forum on Corporate Governance. The mechanism that enforces this is a Long-Term Benefit Trust, or LTBT. The trust holds Class T shares that grant trustees the power to elect a gradually increasing number of directors, starting at one out of five and eventually reaching three enough to constitute a majority of the board, per Harvard Law.
This structure was designed for a company of a different scale. At $1 billion in revenue, the LTBT was a meaningful constraint on investor expectations. At $30 billion and climbing toward $100 billion, the trust holds that same governance weight while the company is navigating defense contracts, international regulatory scrutiny, and infrastructure commitments that would have been unimaginable when the charter was written.
The growth has also changed who is watching. As of October 2025, Anthropic had more than 300,000 business customers accounting for approximately 80% of revenue, according to Sacra. The customer base now spans major financial institutions, healthcare systems, and government-adjacent organizations. The safety mission that defined the company's early positioning is now in direct tension with revenue expectations that require Anthropic to behave like a growth-stage software company rather than a research lab.
Claude Code's market position complicates the picture further. At $2.5 billion in annualized billings and 54% market share in AI programming tools, per Tradingkey, it is Anthropic's most visible product. But the competitive moat Anthropic spent three years building is being partly cannibalized by Claude Code itself, a dynamic Shanaka Anslem Perera identified on his substack as the customer concentration paradox. Claude Code may be expanding the market for AI coding tools while simultaneously compressing the margins on Anthropic's enterprise business.
The $380 billion valuation does not care about any of this. It was set in February 2026 when the growth trajectory was unambiguous and the competitive field was still resolving. Six months later, the competitive field is not resolved. OpenAI is not standing still. DeepMind's models are competitive on key benchmarks. The inference cost curves that Anthropic used to justify the infrastructure buildout are declining faster than the original models assumed, which compresses pricing power across the industry.
The $100 billion question is not a real question. The real questions are whether the LTBT governance mechanism can meaningfully constrain a company that is now larger than most Fortune 500 companies by revenue, whether the gross-basis revenue reporting obscures a margin structure that investors are underwriting at 27x, and whether the safety-first identity that differentiated Anthropic in its early years can survive the commercial logic of $30 billion in annualized revenue.
Anthropic will file its next numbers eventually. The LTBT will have something to say about what they mean.
