Anthropic Just Made History: $10.9B in Q2 Revenue and Its First-Ever Profit in 2026

Anthropic Just Made History No One Expected So Soon

The AI safety company that once told investors it might not turn a profit until 2028 is now on the verge of posting $559 million in operating profit on $10.9 billion in quarterly revenue. That’s not a misprint. That’s Q2 2026 for Anthropic.

According to sources cited by CNBC and Dataconomy, Anthropic has shared projections with investors showing its second quarter of 2026 will be the company’s first-ever operating profit — a seismic shift for an organization that was burning hundreds of millions of dollars per quarter as recently as 2025. The revenue figure of $10.9 billion is more than double the $4.8 billion Anthropic generated in Q1 2026.

Let that sink in: revenue more than doubled in a single quarter. From $4.8 billion to $10.9 billion. In three months.

What’s Actually Driving This Explosive Growth

One product is primarily responsible for this trajectory: Claude Code. Anthropic’s AI coding assistant — which competes directly with OpenAI’s Codex, GitHub Copilot, and a growing array of developer tools — has become the dominant coding AI in enterprise environments.

According to spending data from Ramp, the corporate card and expense management platform, Anthropic has overtaken OpenAI among verified business customers. Enterprise adoption has been the rocket fuel here, not consumer subscriptions.

Consider some of the specific deployments driving revenue:

• Bristol Myers Squibb is making Claude available to more than 30,000 employees for drug discovery workflows — connecting AI directly to pharmaceutical R&D at industrial scale
• Enterprise coding infrastructure: Fortune 500 companies deploying Claude Code across thousands of developer seats, where each seat generates significant recurring revenue
• Government and defense contracts: Multiple federal agencies using Claude for document analysis, code review, and research summarization
• The SpaceX compute deal: Anthropic is paying SpaceX $1.25 billion per month for AI compute access through Colossus 1 and 2 — that’s $41 million per day, confirming the scale of Anthropic’s own infrastructure investment and, indirectly, the revenue it needs to support such costs

Claude’s reputation for accuracy, safety, and literal instruction-following has made it the preferred model for high-stakes enterprise deployments where hallucinations or unexpected behaviors could be costly or dangerous. That’s a moat OpenAI still struggles to match in regulated industries.

The Long Road to This Moment

Anthropic was founded in 2021 by Dario Amodei, Daniela Amodei, and a group of former OpenAI researchers who left over disagreements about safety practices. The company’s central thesis has always been that the most powerful AI systems require the most rigorous safety infrastructure — a bet that has often been mocked as commercially naive.

Last summer, Anthropic told investors it didn’t expect a full-year profit until at least 2028. The turnaround to a profitable quarter in Q2 2026 represents one of the most dramatic reversals in recent tech history — a direct result of enterprise AI spend exploding in 2025-2026 as businesses stopped piloting AI and started deploying it at scale.

The company has raised over $10 billion in funding from investors including Amazon ($4 billion), Google ($2 billion), and a range of strategic partners. Amazon in particular has made Anthropic central to its AI strategy, distributing Claude through Amazon Bedrock to enterprise AWS customers. The Pentagon’s AI deals in early 2026, while excluding Anthropic from some contracts, have since evolved — with the company now part of the critical defense tech ecosystem through Claude’s use in classified research environments.

$559 Million Profit — But Watch the Compute Costs

The $559 million operating profit is significant, but Anthropic watchers should pay close attention to what comes after Q2. The company faces a very specific challenge: its compute costs are scheduled to increase substantially in the second half of 2026 as it scales up training runs for next-generation models.

Running large language models at the quality level of Claude requires massive GPU infrastructure. Anthropic’s deal to pay Colossus $1.25 billion per month for compute access through May 2029 represents a $45 billion commitment — an enormous fixed cost that must be covered by revenue growth continuing at its current pace.

The question isn’t whether Anthropic can sustain this quarter’s profitability — it’s whether it can scale revenue fast enough to stay ahead of compute costs that rise with model capability ambitions. Training Claude 5, when it comes, will not be cheap.

Digitimes reported that Anthropic may face profitability challenges maintaining these margins throughout 2026 due to high scheduled compute costs. The Q2 profit might be the high point before a return to losses — or it might be the beginning of a new financial era for the company. We’ll know by Q3.

How Anthropic Now Compares to OpenAI

For context, OpenAI is now running at roughly $24 billion annualized revenue — still more than double Anthropic’s current quarterly pace. But the gap is narrowing much faster than anyone predicted 18 months ago.

OpenAI has 900 million weekly active users for ChatGPT. Anthropic has a fraction of that consumer base. But Anthropic’s revenue per customer — driven by enterprise contracts and high-value API usage rather than $20/month consumer subscriptions — may actually exceed OpenAI’s on an average basis.

This matters because enterprise revenue is stickier. Consumer subscribers churn. Enterprise contracts renew annually, expand over time, and come with procurement relationships that are difficult to unwind once integrated into workflows. Anthropic may be building a more durable revenue base than its raw user count suggests.

IPO Pressure, Model Releases, and What Comes Next

With OpenAI now filing confidentially for an IPO, the pressure on Anthropic to clarify its own public markets timeline intensifies. The company has been private and mission-driven since founding, but $10.9 billion in quarterly revenue creates new stakeholder expectations.

Amazon and Google — Anthropic’s two largest investors — are public companies with their own shareholders demanding returns. A path to liquidity becomes increasingly important as Anthropic’s valuation climbs.

On the product side, the company continues to iterate rapidly. Claude Mythos — Anthropic’s latest flagship model — has been widely covered, and Claude Code’s dominance in enterprise is expected to deepen as more development shops standardize their AI tooling.

For the broader AI industry, Anthropic’s first profitable quarter is more than a financial milestone. It’s proof that the “AI safety first” approach can be commercially viable — that companies don’t have to choose between building carefully and building successfully. That’s a thesis that has implications for how the entire industry develops going forward.

The company that was told it would never catch OpenAI just recorded $10.9 billion in quarterly revenue. The AI race is far from over — it’s just getting genuinely interesting.