In February 2026 alone, Anthropic added roughly $6 billion to its annualized revenue base. Not for the quarter. Not for the year. For a single month.
The number is hard to absorb. At the start of 2025, Anthropic was a fast-growing AI lab with impressive technology and significant investor backing but still modest commercial scale. By March 2026, Bloomberg reported the company's annualized revenue had surged to approximately $19 billion, more than double what it was generating at the start of the year. The primary driver was Claude Code, Anthropic's coding assistant, which has been adopted at a pace that surprised even the company's internal projections.
Now, Anthropic is considering a next step that would have seemed improbable even a year ago: a public offering, potentially as soon as October 2026, that bankers believe could raise more than $60 billion in new capital.
The Numbers That Got Bankers Calling
The revenue trajectory tells a story that is almost geometrically steep.
| Period | Annualized Revenue | Source |
|---|---|---|
| End of 2025 | $9 billion | Bloomberg, January 2026 |
| February 12, 2026 | $14 billion | Anthropic press release |
| March 3, 2026 | ~$19 billion | Bloomberg |
Each jump represents months of growth that most enterprise software companies achieve only over years. The $5 billion added between the February and March Bloomberg reports took approximately three weeks.
Enterprises account for roughly 80% of Anthropic's business, according to CEO Dario Amodei. CFO Krishna Rao framed the commercial momentum in February: "Whether it is entrepreneurs, startups or the world's largest enterprises, the message from our customers is the same: Claude is increasingly becoming critical to how businesses work."
That statement, dry as it reads in a press release, is backed by the numbers. Something shifted in enterprise adoption in late 2025 and accelerated through early 2026, and the most visible driver has been Claude Code.
Claude Code Changed the Equation
Claude Code, Anthropic's agentic coding product, reached the market at a moment when engineering teams were actively looking for tools that could do more than autocomplete. Unlike earlier coding assistants that operated at the level of individual lines, Claude Code works across entire codebases: running terminal commands, writing and running tests, fixing bugs across multiple files at once, and operating with minimal human supervision.
The adoption curve for developer tools has historically been slow, driven by the friction of changing entrenched workflows. Claude Code bucked that pattern. Paid subscriptions to Claude roughly doubled in 2026, according to an Anthropic spokesperson quoted by TechCrunch in late March. Monthly visits to claude.ai surged from 16 million in January 2025 to 220 million in January 2026, a 13-fold increase in twelve months.
That consumer surge provided a launchpad. But the enterprise contracts (the ones that added $6 billion in a single month) came from a different mechanism: large organizations signing agreements to give their engineering and data teams access to Claude at scale, often through API integrations with internal tools.
For data scientists and ML engineers specifically, Claude Code represents one of the first AI products that has genuinely changed how they work on production systems, not just how they write boilerplate code.
The IPO Calculus
The Information first reported in late March that Anthropic had begun early conversations with investment banks about a potential IPO. Multiple sources confirmed that Goldman Sachs, JPMorgan Chase, and Morgan Stanley are among the institutions in early talks about lead roles in a listing. Bankers familiar with the process, according to reporting from Bloomberg and The Information, expect the offering to raise more than $60 billion if it proceeds.
To put that number in context: the largest IPO by proceeds in recent history was Alibaba's 2014 listing, which raised approximately $25 billion. Anthropic's expected raise, if it materializes, would more than double that record.
The company's most recent private valuation, set in a February 2026 Series G funding round led by GIC, was $380 billion.
A raise of that scale at that valuation would represent roughly 16% of the company going to public shareholders, a standard proportion for a major technology listing.
Anthropic's spokesperson pushed back on the specificity of the reporting: the company "has not decided when or even if it will go public," they said in a statement to Reuters. That kind of careful denial is standard for companies in preliminary IPO discussions. Formal communications lawyers typically advise against confirming anything until SEC filings are required.
The Race Against OpenAI
The timing of the IPO discussion is not accidental. OpenAI has been openly discussing its own path to a public listing, and Anthropic appears to be calculating that an early October window could let it reach public markets before its most prominent rival.
The competitive dynamic has become a recurring theme in 2026. Anthropic's revenue surge has been partly fueled by users who left OpenAI products following the Pentagon deal controversy, a shift tracked in detail in the LDS article Altman Called the Pentagon Deal "Sloppy." 1.5 Million Users Had Already Left. The swing was notable enough that it showed up in Anthropic's commercial metrics within weeks.
Anthropic's own Pentagon situation has created a different kind of headline risk. Defense Secretary Pete Hegseth designated Anthropic a supply-chain risk earlier this year, restricting the company from certain government sales channels. Anthropic subsequently sued the Trump administration, a case covered here: Anthropic Said No to the Pentagon. Now It's Suing the White House.
The company is also developing what it claims is its most powerful model yet, Claude Mythos, details of which leaked inadvertently last week in a data cache incident. Anthropic Accidentally Leaked Its Most Powerful AI Model. Cybersecurity Stocks Crashed the Next Morning.
A Pentagon Ban Did Not Stop the Revenue
The remarkable thing about Anthropic's revenue growth is what it suggests about the revenue's source: it is almost entirely commercial, not governmental.
Most large AI deployments in Washington and the national security apparatus run through OpenAI or smaller specialized vendors. Anthropic's commercial momentum (with enterprises integrating Claude Code across engineering organizations and data teams) has been largely insulated from the Pentagon dispute.
That insulation matters for the IPO calculation. Public market investors assessing an AI company typically want to see revenue that is diversified across industries, not dependent on any single government relationship. Anthropic's current mix, with enterprises at 80% of revenue, provides exactly that kind of story.
The downside risk is regulatory. The company's approach to revenue recognition (specifically, how it accounts for cloud computing credits from major investors like Google and Amazon) has drawn scrutiny. If regulators require Anthropic to restate any figures before a public offering, the timeline could shift significantly.
The Bottom Line
Anthropic is not a company that was supposed to be here yet. In 2023, industry analysts were debating whether it could survive long enough to compete with OpenAI. By March 2026, it has nearly doubled its annualized revenue in a single quarter, secured a valuation that rivals decades-old technology giants, and is in early talks with Wall Street about one of the potentially largest IPOs in history.
The story underneath the headline numbers is simpler than it might appear: engineers and data scientists found a tool that made them demonstrably better at their jobs, paid for it, got their companies to pay for it at enterprise scale, and that enterprise scale turned into $19 billion in annualized revenue in under three months. Whether that trajectory is sustainable (and whether Anthropic can maintain it against a resurgent OpenAI, an aggressive Google, and a rapidly improving open-source field) is the question that a public market will ultimately price.
October is a long time away. The trajectory is not.
Sources
- Anthropic Nears $20 Billion Revenue Run Rate Amid Pentagon Feud, Bloomberg, March 3, 2026
- Anthropic's Revenue Run Rate Tops $9 Billion as VCs Pile In, Bloomberg, January 21, 2026
- Anthropic Eyes $60 Billion IPO as Soon as Q4 2026, WinBuzzer, March 30, 2026
- Anthropic considers IPO as soon as October 2026, Reuters, March 27, 2026
- Anthropic's Claude popularity with paying consumers is skyrocketing, TechCrunch, March 28, 2026
- Enterprises Drive Anthropic Run-Rate Revenue to $19 Billion, PYMNTS, March 2026
- Anthropic ARR surges to $19 billion on Claude Code strength, Yahoo Finance, March 2026