Qwen 3.8 and Claude Opus 5 show why raw benchmark scores don't predict the bill

5 points | by ashurandi 10 hours ago

2 comments