3h agoWeirdML benchmark finds Claude Opus 4.8 trails GPT-5.5 by 2% accuracy while requiring 129 lines of code compared to 517Claude Opus 4.8 costs $2.35 per run versus $2.57.SentimentSentimentPos59.5%Neg40.5%Positive users praise Claude Opus 4.8 for nearly matching GPT-5.5 on WeirdML with far less code and lower future maintenance, while negative users dismiss the results or lament lack of excitement for other models.16 comments with sentiment. View comments.