For a while, the gap between the leading AI labs and everyone else looked wide and fairly stable. That gap appears to be narrowing. Chinese AI lab Moonshot is preparing to release its next model, Kimi 3, which early expectations suggest could close much of the distance with Anthropic’s Opus 4.8, one of the strongest models currently available.
The comparison matters beyond bragging rights. As open and international models get closer to the performance of the best-known US labs, buyers gain more room to negotiate on price and licensing terms rather than defaulting to whichever brand name is most familiar. That’s a meaningful shift for businesses budgeting AI spend, since a shrinking capability gap makes cost, latency, and data handling policies much bigger factors in choosing a model.
It’s also a reminder that no single company’s lead in AI is likely to be permanent. Model quality has moved fast enough in the past two years that any comparison is really a snapshot rather than a settled ranking, so anyone choosing a model for a long-term project should plan to revisit that choice periodically rather than treating today’s leaderboard as fixed.
