logoalt Hacker News

alvisyesterday at 7:10 PM0 repliesview on HN

What surprise me is that Opus 4.5 lost all reasoning scores to Gemini and GPT. I thought it’s the area the model will shine the most