AI 导读
Anthropic 的 Thariq 表示，Claude Code 的“high = 10”输出是数值映射问题，并非暗中降级。 “High”显然仍然是 high，内部评测显示没有性能退化。 感谢澄清。但我还是得说，目前大多数模型用起来都明显变笨了。尤其是 Opus。
正文
Anthropic Thariq says Claude Code’s “high = 10” output was a numerical mapping issue, not a stealth downgrade.
“High” is apparently still high, and internal evals show no performance regression.
Thanks for the clarification. But I still have to say that most models currently feel significantly dumber to use. Especially Opus.
来源：@kimmonismus · x.com