Anthropic's $2 Trillion IPO Hinges on a $190–200 Billion Revenue Forecast

The most consequential number in the AI industry this week is not a benchmark score or a parameter count. It is a revenue forecast. According to an exclusive Reuters report published August 15, Anthropic is projecting annual revenue of roughly $190 billion to $200 billion by 2028, and that single projection is now the fulcrum on which the company’s expected $2 trillion October IPO balances.

Two people familiar with Anthropic’s financials told Reuters that bankers and prospective investors are evaluating the offering using enterprise value-to-revenue multiples anchored to that 2028 forecast. If the number holds up under diligence, the Claude maker’s public debut would be the largest initial public offering in history, surpassing even SpaceX’s record listing. If it does not, the valuation math gets dramatically harder.

From research lab to revenue machine

The trajectory behind that forecast is genuinely unusual. Anthropic generated $4.8 billion in revenue in Q1 2026, according to figures the company shared with investors. In Q2, revenue was on track to more than double to $10.9 billion — a 130% quarter-over-quarter jump that the Wall Street Journal, which first reported the projections in May, described as an “explosive rate of growth.”

More striking than the top line: Q2 was projected to deliver Anthropic’s first-ever quarterly operating profit of roughly $559 million. Very few frontier AI companies have shown profitable operations at any scale, and Anthropic would be the first of the major labs to do so on the back of core product revenue rather than one-off accounting effects.

A significant share of that growth is being driven by Claude Code, Anthropic’s agentic coding tool, which surpassed $1 billion in annualized revenue earlier this year and has become the default entry point for developers adopting the Claude ecosystem. By mid-May, around its $65 billion Series H announcement, Anthropic had disclosed a $47 billion run-rate revenue figure. Reuters now reports the company is on pace for roughly $100 billion in annualized revenue by the end of 2026.

The valuation ladder

Anthropic’s private market valuation has climbed at a pace that has surprised even its own investors. The company’s most recent private round, the Series H announced in late May, valued it at $965 billion — a figure that already put it ahead of OpenAI’s most recent reported valuation of around $852 billion and made it the most valuable AI company in the world.

Now investors briefed on the IPO plans tell the Financial Times and Quartz they expect an October listing at $2 trillion or more. Reuters notes that the implied valuation has climbed some 733% through the valuation ladder’s rungs — a velocity that even optimistic software investors describe as without precedent. Jim Cramer, commenting on the reports, argued that the surging revenue “justifies” the record valuation.

For context on what $2 trillion means: Fortune’s analysis pointed out that an Anthropic valued at that level would need to eventually earn like Amazon — a company that took decades and a dominant cloud franchise to reach comparable profit pools. That is the wager investors are being asked to underwrite: that enterprise AI adoption, agentic coding, and Claude’s enterprise traction compound for years without a competitive reset.

Reasons for skepticism

Not everyone is convinced. Analysts at SaaStr note that on actual calendar-2026 revenue — likely landing somewhere around $20 to $26 billion for the full year — Anthropic would rank in the top five software companies by revenue, not at the very top. A $2 trillion valuation against $20-odd billion of current-year revenue implies a price-to-sales ratio that no large-cap software company has ever sustained, and it effectively prices in flawless execution through 2028.

Critics also caution that the “first operating profit” framing deserves scrutiny: the projection covers operating profit for a single quarter, in a period when Claude Code subscriptions and enterprise contracts are scaling steeply, and it does not include the full burden of compute commitments, stock compensation, and the massive infrastructure buildouts that frontier training runs require. Whether profitability persists across 2027 and 2028 — the years the forecast depends on — remains unproven.

The competitive landscape adds risk. OpenAI is ramping up enterprise growth ahead of its own widely expected IPO, and pricing pressure across the industry is intense: OpenAI cut GPT-5.6 “Luna” pricing by 80% this month, and cheaper Chinese models continue to compress the low end of the market. A study covered by The Information this week even found Anthropic’s models can now be cheaper to run than Chinese alternatives — good for adoption, but a reminder that price competition cuts both ways.

What to watch

The Reuters report gives the clearest picture yet of how the biggest AI IPO in history will be priced: not on current financials, but on a two-year-out revenue forecast of $190–200 billion, discounted by bankers using enterprise value-to-revenue multiples. Three signals will determine whether the $2 trillion target survives to the bell:

1. Q3 revenue continuity. After the projected Q2 jump to $10.9 billion, investors will want to see sequential growth rather than a plateau as the S-1 goes public.
2. Sustained operating profit. One profitable quarter is a milestone; recurring profit is a business model.
3. Claude Code and enterprise retention. With agentic coding the fastest-growing wedge in AI spend, net revenue retention there is the single best proxy for the 2028 forecast’s credibility.

An October listing at $2 trillion would be more than a record — it would be the moment the public markets formally price the AI boom’s second act. The forecast says $200 billion. The market has eight weeks to decide whether to believe it.