That might be your experience. I also prefer Claude for my tasks, but for general usage they are very close. Leaderboards like LLM arena show this and effectiv

That might be your experience. I also prefer Claude for my tasks, but for general usage they are very close. Leaderboards like LLM arena show this and effectively rank all latest models within 20-30 points, which is almost a coin flip. 30 point difference in Elo rating is ~55%/45%, so out of 11 answers, you might prefer 6 from best model, and 5 from worst.