GPT-5 · MMLU Language (0-shot) · OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows. · 成绩与配置

GPT-5 的 MMLU Language (0-shot) · OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows. 记录。本次发布有 28 条成绩记录,含 28 条精确值(可含历史记录),附 56 条来源定位。

成绩与评测条件

GPT-5 · MMLU Language (0-shot) 成绩与配置
指标与成绩配置与评测条件出处
Accuracy90.3%原文 90.3% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Arabic; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy85.7%原文 85.7% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Arabic; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy89.2%原文 89.2% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Bengali; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy85%原文 85.0% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Bengali; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy90.2%原文 90.2% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Chinese (Simplified); translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy86.7%原文 86.7% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Chinese (Simplified); translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy90.1%原文 90.1% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), French; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy87.5%原文 87.5% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), French; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy89.6%原文 89.6% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), German; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy86.6%原文 86.6% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), German; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy89.9%原文 89.9% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Hindi; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy86.1%原文 86.1% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Hindi; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy90.9%原文 90.9% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Indonesian; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy87.2%原文 87.2% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Indonesian; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy90.8%原文 90.8% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Italian; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy87.6%原文 87.6% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Italian; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy89.8%原文 89.8% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Japanese; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy86.5%原文 86.5% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Japanese; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy89.6%原文 89.6% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Korean; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy85.4%原文 85.4% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Korean; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy91%原文 91.0% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Portuguese (Brazil); translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy87.9%原文 87.9% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Portuguese (Brazil); translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy91%原文 91.0% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Spanish; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy88.1%原文 88.1% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Spanish; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy88%原文 88.0% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Swahili; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy81.5%原文 81.5% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Swahili; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy80.6%原文 80.6% · 单位 %
公布
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Yoruba; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩
Accuracy66.4%原文 66.4% · 单位 %
公布
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MMLU Language (0-shot), Yoruba; translated test set; chain-of-thought prompting.
条件
Table 10 lists this language; Section 3.10 says 13 languages, while the table lists 14; exact reasoning effort is not reported.
在排名中查看此成绩

出处与披露

测试版本与指标

别名 MMLU Language (0-shot; GPT-5 System Card)

  • Accuracy · % · 越高越好