Developer traffic to OpenAI and Anthropic has fallen, even as both companies roll out more capable AI models. The decline marks a shift for the two firms, which have been central to the boom in generative AI tools over the past two years.
The drop in developer engagement could complicate their efforts to maintain dominance in the fast-moving AI market. It also raises questions about whether model improvements alone are enough to keep coders and startups coming back.
What the traffic dip looks like
OpenAI and Anthropic don't publish detailed metrics on developer activity. But the contraction is real enough to be noted across the industry. Developers who once flocked to their APIs for prototyping and production workloads are either scaling back or spreading their attention across a wider set of providers.
The reasons aren't fully clear. Some of it may be natural consolidation as the initial wave of AI experimentation gives way to more deliberate adoption. Some may stem from cost pressures or technical friction. What's certain is that the audience both companies spent years courting is no longer growing the way it did.
Developer traffic isn't just a vanity metric. It's a leading indicator of platform stickiness. When developers build on top of a model, they tend to stay—unless a rival offers something meaningfully better, cheaper, or easier to use. A sustained decline can erode the network effects that make a platform hard to leave.
Both companies have raised billions on the premise that they'll capture the lion's share of the AI application layer. If developers start drifting, that thesis gets harder to defend. It also gives ammunition to competitors—whether open-source alternatives or cloud giants with their own model portfolios—who argue that no single provider should be a default.
Model improvements haven't reversed the trend
Over the past year, OpenAI and Anthropic have shipped faster, cheaper, and more capable models. On paper, those releases should attract more developers, not fewer. But better benchmarks don't automatically translate into deeper engagement. Developers care about reliability, latency, pricing, and how well a model fits into their existing stack. They also care about not getting locked in.
The fact that traffic has slipped despite technical progress suggests the competitive landscape is shifting. It's no longer enough to have the best model. Distribution, tooling, and trust matter just as much—and rivals are catching up on all three.
Neither OpenAI nor Anthropic has publicly addressed the developer traffic decline. Both are likely watching the numbers closely, especially as they prepare for new model launches and enterprise pushes. The next few quarters will show whether this is a blip or the start of a longer-term cooling.
For now, the pressure is on to convert model upgrades into renewed developer interest. If that doesn't happen, the two companies may need to rethink how they win back the builders who once made them the default choice.
