Peter Yang 实测 Claude Fable 5.2 与 Opus 5.5：五个用例中 Fable 胜四场
Fable 5.2 is here! I had early access and tested @claudeai's new model head to head against Opus 5.5 across 5 use cases:
Peter Yang 提前拿到 Claude Fable 5.2，与 Opus 5.5 在五个用例上做了对比测试，包括日本旅行视频、3D 卡丁车游戏、网站改版、写一篇 600 词不注水的文章，以及在他提出反驳时坚持立场。Fable 在五场中赢了四场，Opus 仍赢下一场，具体是哪一场需看他的视频。Claude 官方同日宣布推出 Claude Fable 5.2 和 Claude Mythos 5.2，称两者在编码、推理和视觉等方面有所提升。
推荐理由
AI 评分 58
作者用五个具体用例对比两款新模型，给出胜负分布，可据此了解各自擅长的任务类型。
主题
查看原帖隆重推出 Claude Fable 5.2 和 Claude Mythos 5.2，它们旨在攻克最艰巨的技术挑战。 在许多领域，它们都是雄心勃勃项目中的得力协作者，并在编码、推理和视觉等方面展现出提升。https://t.co/jP0AutLxXb
查看原文
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges. They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision. https://t.co/jP0AutLxXb
正文
Fable 5.2 is here! I had early access and tested @claudeai's new model head to head against Opus 5.5 across 5 use cases:
1. Travel video of Japan 2. 3D kart racing game 3. Website redesign 4. Writing a 600-word essay without slop 5. Holding its ground when I push back
Fable won 4 of 5, but Opus still won one. Watch my video to find out which one.
📌 Watch here: https://t.co/XbKPhJ8Dwq