Claude Opus 4.7 · DRACO(100-task Perplexity deep-research benchmark) 77.7% · 附件证据(ev-op47-card-draco)

Claude Opus 4.7 的 DRACO · 100-task Perplexity deep-research benchmark 官方出处。DRACO,max effort · 77.7%。见《Claude Opus 4.7 System Card》§8.8.4 and Figure 8.8.4.A, PDF pp.201-202。证据 ev-op47-card-draco。

原文与定位

Claude Opus 4.7 · DRACO(100-task Perplexity deep-research benchmark),DRACO,max effort配置,Normalized score 77.7%(具体成绩公布时间未确认)。见《Claude Opus 4.7 System Card》§8.8.4 and Figure 8.8.4.A, PDF pp.201-202。 查看成绩条件

Opus 4.7 scored 77.7% with adaptive thinking at max effort and a 1M token limit. We used context compaction (triggered at 200k tokens).
官方材料
Claude Opus 4.7 System Card · PDF 第 201 页
定位
附件 · §8.8.4 and Figure 8.8.4.A, PDF pp.201-202
材料版本
Current saved PDF/TXT: 本地留存副本 and current-opus-4-7-system-card.txt; 232 PDF pages; metadata created/modified 2026-04-16 22:11:48 China Standard Time. Full current-card review completed 2026-11-16; this file is not established as launch-day bytes.
版本公布
未确认
披露时间
未确认 · 首发/后补未确认
采集时间
历史依据
暂无可靠历史版本依据
引用文本

Claude Opus 4.7 · DRACO(100-task Perplexity deep-research benchmark),DRACO,max effort配置,Normalized score 77.7%(具体成绩公布时间未确认)。官方出处 Claude Opus 4.7 System Card(附件,Current saved PDF/TXT: 本地留存副本 and current-opus-4-7-system-card.txt; 232 PDF pages; metadata created/modified 2026-04-16 22:11:48 China Standard Time. Full current-card review completed 2026-11-16; this file is not established as launch-day bytes.)§8.8.4 and Figure 8.8.4.A, PDF pp.201-202,https://www.anthropic.com/claude-opus-4-7-system-card#page=201。Benchlog https://benchlog.lekshome.cc/evidence/claude-opus-4-7/ev-op47-card-draco/

关联成绩与条件

Claude Opus 4.7 · DRACO 成绩与配置
指标与成绩配置与评测条件证据关系
Normalized score77.7%原文 77.7% · 单位 %
公布 未确认
DRACO,max effort
设置
100 tasks; web search, web fetch, programmatic tool calling and code execution; adaptive thinking, max effort; 1M token limit; context compaction at 200K; five independent Opus 4.6 grading runs per response, mean normalized score.
未知项
未另记
工具
允许工具
推理预算
max
框架
未记录
协议
DRACO; normalized score over 100 curated research tasks.
条件
Adaptive thinking, max effort; web search, web fetch, programmatic tools and code execution; 1M token limit; compaction at 200K; single model run averaged over five independent Opus 4.6 grading runs.

直接支持该成绩

在排名中查看此成绩