用户吐槽 Opus 5 质量下滑,称 GPT-5.6 远胜

Chubby♨️ · @kimmonismus · X·2026-08-03 21:03·17天前
AI 导读

Kim 称 Opus 5 是首个越用越不喜欢的模型,常偏离指令、遗忘任务且受护栏限制,认为 GPT-5.6 远胜。引用推文指 Opus 5 是 Anthropic 迄今最差模型,频繁索要批准并擅自行动。Kim 还批评 Sonnet 5 发布不佳,且自 Opus 4.6 后模型质量明显下滑。

Chubby♨️@kimmonismus
29AI 编辑部评分,满分 100

用户吐槽 Opus 5 质量下滑,称 GPT-5.6 远胜

2026-08-03 21:03· 17天前
AI 导读

Kim 称 Opus 5 是首个越用越不喜欢的模型,常偏离指令、遗忘任务且受护栏限制,认为 GPT-5.6 远胜。引用推文指 Opus 5 是 Anthropic 迄今最差模型,频繁索要批准并擅自行动。Kim 还批评 Sonnet 5 发布不佳,且自 Opus 4.6 后模型质量明显下滑。

i thought it was just me. Opus 5 is the first model that the more I use, the worse I like it.

It goes off-track, does random things I didn't ask for, forgets other things, and constantly has those annoying guardrails, etc. Sorry, but GPT-5.6 is worlds better.

I really don't understand it anymore. What's going on at Anthropic? Sonnet 5 was already a bad release, and since Opus 4.6, I feel like the subsequent Opus models have simply declined significantly in quality. My enthusiasm for Opus 5 has definitely evaporated.

🥭Opus 5 is actually worse than any model anthropic has released thus far. Nothing comes close to how much this shit model constantly wants your approval, how it ...

来源:Chubby♨️· x.com