Unverified70% confidenceEventTime unknown
一位名为Evand的评测作者用GPT 5.5重新测试了两周Codex，结论是Codex在很多任务上跟Claude Code已经是一个量级，某些场景下更快更准
1
Sources
70%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026
Sources
Codex vs Claude Code：半年后我为什么又换回来了
bilibiliAli厂长
Related Entities
Related Claims
UnverifiedGPT-4.2 Codex在专业编码基准测试上领先77% similarVerifiedCodex此前落后于Claude Code，但随着GPT最新版本发布和内部优化，其能力已大幅提升77% similarUnverifiedAntiGravity's document reading and code generation speed far exceeded GPT-5.2 in Codex during the refactoring test75% similarVerifiedGPT-5 Codex在复杂任务（前10%）上会使用两倍以上的Token来进行更深入的推理75% similarVerifiedOpenAI的Codex配合GPT-5据称编程能力可与Claude Code媲美74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/13060API
curl https://kongchang.com/api/v1/knowledge/claims/13060MCP
get_claim(id=13060)