Mras (@xieyang)为什么感觉 gpt 5.2 high 模型还是不如 claude ops 呢 中发帖

要求找 android 开发中的 bug,gpt 5.2 medium 和 high 都定位不到,ops  4.5很快就找到了 ,是我的错觉吗 🫪