Unverified60% confidenceBenchmarkExact time
The original Codex achieved a pass@1 score of 28.8% on the HumanEval benchmark, whereas using GPT-3 directly yielded only 0%.
2
Sources
60%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
国内直连Codex配置教程：GPT Agent完整上手指南
bilibili无汁西瓜6/26/2026
Related Claims
UnverifiedCodex于2021年发布时在HumanEval基准上达到28.8%的pass@1得分，而GPT-4在同一基准上已超过67%83% similarVerifiedCodex achieved a pass@1 accuracy of 28.8% on the HumanEval benchmark76% similarUnverifiedThe strongest configuration, GPT-5.5 paired with Codex, achieved a pass rate of only 28.0% on Harbor-Index74% similarUnverified初版 Codex 采用约120亿参数规模，在 HumanEval 基准测试中达到 28.8% 的 pass@1 通过率68% similarUnverifiedIn a certain programming benchmark, OpenAI's GPT-5.5 scored 76% in the benchmark's built-in Harness, but only 57% within Codex internally.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/488714API
curl https://kongchang.com/api/v1/knowledge/claims/488714MCP
get_claim(id=488714)