Domestic models are basically at the level of “keeping up” right now, with occasional bright spots. But if you say they’ve “caught up,” GPT-5.6 is right over there—go and compare them in practice.
Domestic models are basically at the level of “keeping up” right now, with occasional bright spots. But if you say they’ve “caught up,” GPT-5.6 is right over there—go and compare them in practice.