Anthropic Eyes First Profit on $10.9B Q2 Revenue, Two Years Early

Anthropic told investors this week that it expects to post $10.9 billion in revenue and a $559 million operating profit in Q2 2026, putting the AI lab on track for its first-ever profitable quarter — two years ahead of management's earlier guidance.

The figures, first reported by The Wall Street Journal and confirmed by CNBC, were shared as part of an ongoing funding round that could value Anthropic at $900 billion. If the company hits the target, Q2 revenue alone will exceed Anthropic's total revenue for all of 2025.

The Numbers

Anthropic's quarterly trajectory tells the story:

Q1 2026 revenue came in at $4.8 billion. Q2 2026 is projected at $10.9 billion — a 127% sequential jump. The implied annualized run rate is over $43 billion, putting Anthropic in the same revenue tier as enterprise software companies that took 15+ years to scale to that level.

The $559 million projected operating profit is the headline. As recently as last summer, Anthropic was telling investors it did not expect full-year profitability until at least 2028. A profitable Q2 in 2026 pulls that timeline forward by roughly two years.

Why this matters: Anthropic becomes the first frontier AI lab to demonstrate that the commercial model — large enterprise contracts, API access, and a fast-growing consumer product — can actually pay for the underlying compute. That's the central open question hanging over the entire AI industry, and Anthropic is the first to answer it with concrete numbers.

What's Driving It

Three businesses are powering the surge. Anthropic's enterprise API business, particularly through AWS Bedrock and Google Cloud, has scaled rapidly as Fortune 500 companies have moved Claude into production workloads. Direct API contracts with large customers — Microsoft, Salesforce, and a long list of financial services firms — added meaningful revenue in Q1.

Claude Code, Anthropic's coding-agent product launched in 2025, has become one of the fastest-growing developer tools in the industry. The company has not disclosed seat counts, but multiple analysts estimate Claude Code revenue at roughly $1 billion in annualized recurring revenue exiting Q2.

The third leg is the consumer business. Claude.ai's paid subscription tiers have grown alongside the launch of Claude Design, Claude for Small Business, and ad-free Claude in select markets. None of those products alone moves the needle, but in aggregate they contribute meaningfully to the gross margin profile.

The Important Caveat

The same investor disclosure that revealed the Q2 numbers also included a sharp warning: Anthropic may not stay profitable through the rest of 2026.

The reason is on the cost side. Anthropic just signed a $45 billion compute contract with SpaceX for access to the Colossus supercomputing cluster in Memphis, paying $1.25 billion per month through May 2029. That deal, revealed in SpaceX's S-1 IPO filing on May 20, adds roughly $3.75 billion per quarter to Anthropic's cost base — a number that arrives with full force starting in July when discounted ramp-up pricing ends.

The company has also disclosed massive multi-year commitments to Google Cloud (for TPU access) and Amazon Web Services (for Trainium-based training). Total compute commitments through 2029 now exceed $200 billion across all three clouds.

In other words: the revenue is growing fast, but the cost base is growing faster in absolute terms. Whether Q2 profitability is a one-quarter milestone or the start of a sustained trend depends entirely on whether the revenue curve can keep up.

Compared to OpenAI

The disclosure inevitably draws comparisons to OpenAI, which is on track for an even larger annual revenue base but has not yet projected a profitable quarter. OpenAI was last reported targeting profitability in 2029.

The valuation math is now genuinely close. Anthropic's ongoing funding round implies a $900 billion valuation, while OpenAI's most recent secondary tender priced at roughly $850 billion. If Anthropic's round closes at the upper end and OpenAI's flat secondary holds, Anthropic would surpass OpenAI on valuation for the first time.

That would be a remarkable inversion. As recently as 2024, OpenAI traded at roughly three times Anthropic's valuation. The gap closed in 2025, and Anthropic's faster path to profitability appears to be tipping it in the lead heading into both companies' expected IPOs.

Growth Rate Context

The investor materials reportedly framed Anthropic's quarterly growth in historical terms. The company's Q1-to-Q2 jump is faster than Zoom during the pandemic and faster than Google's and Facebook's growth rates in the quarters before their IPOs.

That framing is striking but also worth contextualizing. Zoom's pandemic surge happened in a much smaller absolute revenue base. Google and Facebook were growing into much earlier-stage advertising businesses. Anthropic is growing this fast on a base that is already in the multi-billion-dollar range — which is, if anything, more impressive than the historical comparisons suggest.

What's Next

Anthropic's funding round is expected to close in the next several weeks at the $900 billion valuation. Investors have reportedly oversubscribed the round, with allocations being cut for late-arriving funds.

After that, attention shifts to a possible IPO. Anthropic has not committed to a public listing in 2026, but multiple investors have said the company is eyeing one this year — and a profitable quarter is exactly the kind of milestone that makes a public-market debut viable.

For competitors and customers, the message is that the AI cost-revenue puzzle is solvable. Anthropic is the first to show it, but won't be the last. Expect OpenAI and Google DeepMind to point to their own paths to profitability in the coming months.

The Bottom Line

Anthropic just pulled its profitability timeline forward by two full years on a revenue base growing 127% quarter-over-quarter. The caveats are real — compute costs are about to scale up sharply — but the underlying milestone matters. The question for the AI industry has shifted from "can this be a real business?" to "how big can the real business get?"