Claude Opus 5 · DeepSearchQA · 900 prompts;单代理与多代理按协议区分 · 成绩与配置

Claude Opus 5 的 DeepSearchQA · 900 prompts;单代理与多代理按协议区分 记录。本次发布有 5 条成绩记录,含 5 条精确值(可含历史记录),附 1 条来源定位。

成绩与评测条件

Claude Opus 5 · DeepSearchQA 成绩与配置
指标与成绩配置与评测条件出处
F1 score91%原文 91% · 单位 %
公布 未确认
DeepSearchQA · Low effort
设置
900 multi-hop web-research prompts; web search, web fetch, and programmatic tool calling; adaptive thinking; figure states a 980K-token budget; section states a 1M budget and no context compaction. Explicit adaptive-thinking effort: low.
未知项
The card does not state any additional sampling parameters beyond the cited evaluation description.
工具
允许工具
推理预算
low
框架
未记录
协议
DeepSearchQA; mean F1 across 900 multi-hop web-research questions.
条件
Explicit effort labels; figure budget 980K tokens; section describes web search, web fetch, programmatic tools, 1M token budget, no context compaction.
在排名中查看此成绩
F1 score92.8%原文 92.8% · 单位 %
公布 未确认
DeepSearchQA · Medium effort
设置
900 multi-hop web-research prompts; web search, web fetch, and programmatic tool calling; adaptive thinking; figure states a 980K-token budget; section states a 1M budget and no context compaction. Explicit adaptive-thinking effort: medium.
未知项
The card does not state any additional sampling parameters beyond the cited evaluation description.
工具
允许工具
推理预算
medium
框架
未记录
协议
DeepSearchQA; mean F1 across 900 multi-hop web-research questions.
条件
Explicit effort labels; figure budget 980K tokens; section describes web search, web fetch, programmatic tools, 1M token budget, no context compaction.
在排名中查看此成绩
F1 score94.1%原文 94.1% · 单位 %
公布 未确认
DeepSearchQA · High effort
设置
900 multi-hop web-research prompts; web search, web fetch, and programmatic tool calling; adaptive thinking; figure states a 980K-token budget; section states a 1M budget and no context compaction. Explicit adaptive-thinking effort: high.
未知项
The card does not state any additional sampling parameters beyond the cited evaluation description.
工具
允许工具
推理预算
high
框架
未记录
协议
DeepSearchQA; mean F1 across 900 multi-hop web-research questions.
条件
Explicit effort labels; figure budget 980K tokens; section describes web search, web fetch, programmatic tools, 1M token budget, no context compaction.
在排名中查看此成绩
F1 score94.6%原文 94.6% · 单位 %
公布 未确认
DeepSearchQA · Xhigh effort
设置
900 multi-hop web-research prompts; web search, web fetch, and programmatic tool calling; adaptive thinking; figure states a 980K-token budget; section states a 1M budget and no context compaction. Explicit adaptive-thinking effort: xhigh.
未知项
The card does not state any additional sampling parameters beyond the cited evaluation description.
工具
允许工具
推理预算
xhigh
框架
未记录
协议
DeepSearchQA; mean F1 across 900 multi-hop web-research questions.
条件
Explicit effort labels; figure budget 980K tokens; section describes web search, web fetch, programmatic tools, 1M token budget, no context compaction.
在排名中查看此成绩
F1 score95%原文 95% · 单位 %
公布 未确认
DeepSearchQA · Max effort
设置
900 multi-hop web-research prompts; web search, web fetch, and programmatic tool calling; adaptive thinking; figure states a 980K-token budget; section states a 1M budget and no context compaction. Explicit adaptive-thinking effort: max.
未知项
The card does not state any additional sampling parameters beyond the cited evaluation description.
工具
允许工具
推理预算
max
框架
未记录
协议
DeepSearchQA; mean F1 across 900 multi-hop web-research questions.
条件
Explicit effort labels; figure budget 980K tokens; section describes web search, web fetch, programmatic tools, 1M token budget, no context compaction.
在排名中查看此成绩

出处与披露

测试版本与指标

别名 DeepSearchQA、DeepSearchQA with multi-agents

  • F1 score · % · 越高越好
  • Accuracy · % · 越高越好