Claude Just Lapped ChatGPT and Nobody Seems That Surprised

There’s a particular moment in a race where the person who’s been comfortably in front realises they’re not anymore. They don’t necessarily slow down. Someone else just got faster, and they were looking the wrong way when it happened.

That’s roughly where we are with ChatGPT and Claude.

The numbers doing the rounds this week are genuinely striking. Anthropic’s annualised revenue run rate hit $30 billion in early April, ahead of OpenAI’s $24 to $25 billion at the same point. More U.S. businesses paid for Claude than ChatGPT in April, apparently for the first time ever. Eight of the Fortune 10 are now Claude customers. That last one is the one I keep coming back to.

Worth noting: annualised run rate is a projection, not a final ledger. One strong month multiplied by twelve can flatter or mislead depending on timing. Someone in a comment thread pointed this out, and they’re right. But even discounting for that, the direction of travel here is pretty clear.

I’ve been using Claude more heavily myself over the past few months, mostly for work, some writing, occasionally helping me reason through a gnarly bit of infrastructure logic. The difference in how it responds is noticeable in a way that’s hard to pin down at first and then becomes obvious. It doesn’t pad. It doesn’t produce thirty bullet points in response to a simple question. It seems to actually read what you wrote rather than pattern-matching to the nearest FAQ. Whether that’s the context window, instruction-following, or something else in the training, I don’t know. Probably all of it.

The enterprise shift makes a certain kind of sense when you think about what businesses actually need from these tools. Consumer products can get away with being impressive and occasionally wrong. Enterprise tooling, especially anything touching code or workflow automation, needs to be reliable in a boring way. It needs to do what you told it to do. ChatGPT has always felt, to me at least, like a product designed around the demo. Claude feels more like a product designed around the second hour of use.

The agentic coding angle is where things get genuinely interesting from a market perspective. A few people in various discussions have made the point that OpenAI has been chasing existing markets: search, shopping, social. Anthropic has been building out specific workflows, especially around coding agents, where the value proposition is concrete and measurable. Enterprise procurement teams don’t buy vibes. They buy tools that solve a specific problem and can be justified in a budget meeting. “Our dev team ships faster” is a much easier sell than “it’s a really impressive chatbot.”

OpenAI isn’t dead, obviously. They still have enormous consumer mindshare, GPT-4o is genuinely good, and they have more resources than most countries. But there’s something interesting happening at the top of the market, and I think the honest read is that Anthropic made deliberate choices about where to focus and those choices are now showing up in the revenue line.

I’ll admit I find myself more curious about what comes next than worried about who’s currently winning. The pace of this stuff is still genuinely unsettling in a way I haven’t fully processed. A company goes from $9 billion annualised to $30 billion in roughly four months. That’s not a market maturing. That’s something else, and I’m not sure we have good language for it yet.

What I do know is that “ChatGPT” has become one of those brand names that stands in for an entire category, the way “Google” did for search. That’s a real advantage and also a kind of trap. When the category moves, the brand can get stuck.