Here’s something you don’t hear every day: the CEO of the most hyped AI company on the planet openly admitting a competitor beat them to the punch. Sam Altman, in a March 2026 interview with Wired, conceded that OpenAI fell behind Anthropic’s Claude Code in the AI coding tools race.
The numbers tell the story pretty clearly. Anthropic’s Claude Code was generating nearly $2.5 billion in annualized revenue by February 2026, representing about 20% of the company’s overall business. OpenAI’s competing product, Codex, reported just over $1 billion in annualized revenue at the end of January 2026.
First to market is worth a lot
Altman’s candid assessment carried a particularly ironic edge. “First to market is worth a lot. We had that with ChatGPT,” he told Wired. The implication is hard to miss: OpenAI knows exactly how powerful first-mover advantage can be because they rode that wave themselves when ChatGPT exploded onto the scene.
The difference is that OpenAI fumbled the coding tools opportunity. The company actually had early roots in this space, launching the original Codex back in 2021. But after ChatGPT became a cultural phenomenon, OpenAI’s attention shifted toward broader AI model development rather than dedicated coding agents.
Anthropic, meanwhile, went all-in on developer tools. And by spring 2026, that bet had paid off handsomely, with Claude Code helping Anthropic overtake OpenAI in US enterprise revenue.
The Windsurf deal that wasn’t
OpenAI didn’t ignore the problem entirely. They tried to buy their way back into the game. In 2025, the company explored a roughly $3 billion acquisition of Windsurf, an AI coding startup that could have instantly bolstered OpenAI’s capabilities in the developer tools space.
The deal fell apart. Partnership complexities with Microsoft, OpenAI’s primary backer and distribution partner, created complications that ultimately killed the transaction.
Without Windsurf, OpenAI was left to build its way out of the hole organically. The company has since released GPT-5.5, aimed at closing performance gaps on industry benchmarks.
What this means for the AI investment landscape
Anthropic’s success with Claude Code demonstrates something investors should pay close attention to: in AI, the company with the best foundational model doesn’t necessarily win the most revenue. Anthropic carved out a specific, high-value niche and dominated it while OpenAI was trying to be everything to everyone.
For anyone watching the private market valuations of these companies, the coding tools revenue split is a critical data point. Anthropic overtaking OpenAI in US enterprise revenue by spring 2026 represents a meaningful shift in how these companies should be assessed, given that enterprise revenue tends to be stickier and more predictable than consumer-facing revenue.
The public sparring between the two companies has already intensified, with competing advertisements and pointed statements about business models.