返回首页
AI on Medium··行业媒体

I Tested GPT-5.5 vs Claude Opus 4.8 for 2 Weeks on My Real Work. Forget the Benchmarks.

中文摘要

GPT-5.5对比Claude Opus 4.8,真实工作两周实测,绕过基准测试。

English Summary

GPT-5.5 vs Claude Opus 4.8, tested on real work for 2 weeks, bypassing benchmarks.

原文节选

Here is what actually happened when I stopped reading leaderboards and started testing. Continue reading on Medium »