Fortune confirmed on 2 July that Anthropic has overtaken OpenAI in self-reported annualised revenue — a milestone that seemed unthinkable eighteen months ago when OpenAI held a commanding lead in both consumer adoption and enterprise spending.
Anthropic reported it was on course to hit $47 billion in annualised revenue as of May 2026, while OpenAI stated it would generate between $25 billion and $33 billion in its most recent disclosure. The gap is substantial: Anthropic's run rate is roughly 50 to 90 per cent higher than its rival's, depending on where OpenAI's actual figure falls within its stated range.
Ramp data cited by Fortune shows Anthropic overtook OpenAI in business subscriptions in May 2026 — a leading indicator that enterprise customers are voting with their wallets. The shift is driven by Claude's dominance in coding, agentic workflows and enterprise integration, areas where Anthropic has consistently outperformed since the launch of Claude Opus 4 in late 2025.
On the consumer side, Similarweb data indicates ChatGPT's monthly visits fell below 50 per cent of the generative AI market for the first time. The decline reflects a combination of Claude's growing consumer adoption, the rise of Chinese alternatives like DeepSeek and Xiaomi MiMo, and Google's Gemini capturing search-integrated AI traffic.
The revenue reversal accompanies a valuation shift. Anthropic's $65 billion fundraising round in Q2 2026 elevated its post-money valuation to $965 billion — surpassing OpenAI for the first time and positioning it as the world's most valuable AI startup. Anthropic has also projected profitability by 2029, a year ahead of OpenAI's disclosed timeline, having already reported profitability in Q4 2025.
The divergence reflects fundamentally different strategic bets. Anthropic focused early on enterprise customers using Claude for high-value tasks — customer service, content analysis, code generation and compliance workflows — building large, recurring contract revenue. OpenAI bet on consumer subscriptions and platform breadth, a strategy that generated massive user numbers but lower revenue per user.
Sam Altman acknowledged the competitive pressure in a Fortune interview, describing a 'new world order for AI' in which OpenAI can no longer assume market dominance. The company's response includes the three-tier GPT-5.6 model family (Sol, Terra, Luna), the Jalapeño custom inference chip, and a push into hardware via the Jony Ive acquisition — a diversification strategy that broadens OpenAI's surface area but diffuses its focus.
For context engineers, the revenue flip is a validation of the 'build the best model' strategy over the 'build the biggest platform' strategy. Anthropic's lead in coding and agentic tasks — the workloads that generate the most enterprise value — translated directly into revenue dominance. Developers choosing which models to build on now have clear market signals about where the industry's centre of gravity is shifting.
