Anthropic Revenue Hits $45B ARR, Surpasses OpenAI by 35% Ahead of IPOs

Anthropic’s annualized recurring revenue has reached nearly $45 billion, pushing it roughly 35% past OpenAI’s estimated $33 billion ARR, according to Sherwood News, citing reporting from The Information. The milestone marks a dramatic reversal from even six months ago, when OpenAI was widely considered the runaway leader in commercial AI.

Anthropic’s revenue has quintupled from $9 billion at the end of 2025. The growth rate — roughly 10x per year since hitting the $1 billion ARR milestone — far outpaces OpenAI’s 3.4x annual growth from the same starting point, according to analysis from Epoch AI. Both companies project slower growth in 2026, but the crossover has already happened.

Anthropic is projecting its first‑ever quarterly operating profit of $559 million in Q2 2026, on revenue of at least $10.9 billion — more than double the $4.8 billion it posted in Q1, Reuters reported. The contrast with OpenAI is stark: OpenAI projects $14 billion in losses for 2026 and has pushed its breakeven target to approximately 2029.

This is not just a headline number. Anthropic’s path to profitability — three years ahead of OpenAI’s timeline — changes the math for investors sizing up the dueling IPOs. Anthropic is expected to reach positive free cash flow by 2027, while OpenAI continues to spend roughly 4x more on compute to train models that generate less revenue, according to industry analysis. OpenAI’s compute spending is projected to reach $121 billion in a single year.

The revenue gap is explained by where the money comes from. 80% of Anthropic’s revenue is from enterprise customers, with over 1,000 businesses each spending more than $1 million annually — a number that doubled in under two months. These are sticky, expanding contracts, not consumer subscriptions that churn.

OpenAI’s revenue mix leans more heavily on converting individual users from free to paid tiers. Anthropic’s product lineup — Claude Cowork for team workflows, Claude Code for engineering teams, and the Claude + Microsoft 365 connector — replaced line items in enterprise budgets, not just added new ones. Claude is also the only frontier model available on all three major cloud platforms: AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry. OpenAI is primarily on Azure, though its exclusivity agreement has ended and the company is moving toward multi‑cloud support.

Anthropic’s 2026 release cadence has been unusually aggressive, with major releases roughly every two weeks since January. Key products that drove enterprise adoption include Claude Cowork, which triggered what one analyst described as a $285 billion software selloff by replacing entire team workflows, and Claude Code, which became a de facto engineering chief‑of‑staff at many organizations.

The Mythos security model, which finds software vulnerabilities at machine speed, has also been a significant enterprise draw. The model has already uncovered thousands of high‑severity flaws across major operating systems and browsers, per Anthropic’s disclosures, and has been adopted through the Project Glasswing program by companies including Zscaler and government agencies.

Both Anthropic and OpenAI are targeting IPOs later this year, alongside SpaceX’s $75 billion offering, as reported by HPC Wire. The trio of offerings would mark the first opportunity for public investors to directly participate in the AI boom, and the financial disclosures from Anthropic and OpenAI will provide groundbreaking transparency into the economics of frontier AI.

Anthropic’s revenue lead and profitability timeline give it a stronger story for public markets. But OpenAI retains brand recognition and breadth — its Q1 2026 quarterly revenue of $5.7 billion still led Anthropic’s $4.8 billion in absolute quarterly terms, per Sherwood News. The annualized figures tell a different story because of Anthropic’s steeper growth curve, but the quarterly battle is still close.

The revenue dynamics have practical implications for developers choosing AI platforms. Anthropic’s multi‑cloud availability means builders can use Claude through their existing cloud provider relationships — AWS, Google Cloud, or Azure — without adding new vendor contracts. OpenAI’s Azure‑only distribution limits that flexibility.

Anthropic’s profitability also suggests pricing stability: a company approaching free cash flow positive has less pressure to raise API prices to cover losses. OpenAI’s $14 billion projected loss for 2026 could lead to price increases or feature gating. For builders making long‑term bets on AI infrastructure, the financial health of the underlying lab matters as much as benchmark scores.