Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark) · 附件证据(ev-f51-card-draco)
Claude Fable 5.1 的 DRACO · 100-task Perplexity deep-research benchmark 官方出处。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。证据 ev-f51-card-draco。
Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Low effort配置,Normalized score 84.6%(具体成绩公布时间未确认)。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。 查看成绩条件
Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Medium effort配置,Normalized score 85.3%(具体成绩公布时间未确认)。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。 查看成绩条件
Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),High effort配置,Normalized score 85.4%(具体成绩公布时间未确认)。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。 查看成绩条件
Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Xhigh effort配置,Normalized score 87.1%(具体成绩公布时间未确认)。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。 查看成绩条件
Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Adaptive Max配置,Normalized score 87.7%(具体成绩公布时间未确认)。见《Claude Fable 5.1 and Claude Mythos 5.1 System Card》本地留存副本, PDF p.179。 查看成绩条件
Figure 8.12.2.A: Claude Fable 5.1 DRACO normalized scores by effort: low 84.6%, medium 85.3%, high 85.4%, xhigh 87.1%, max 87.7%.
PDF dated September 1, 2026; 212 pages. Full text review completed: pp.1-212. All 64 low-text or figure pages were rendered and visually inspected in 16 sheets; no additional standard Fable 5.1 capability scores were found in the supplementary safety/alignment sections. Revision date unknown. Local: 本地留存副本 and .txt.
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 Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Low effort配置,Normalized score 84.6%(具体成绩公布时间未确认);Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Medium effort配置,Normalized score 85.3%(具体成绩公布时间未确认);Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),High effort配置,Normalized score 85.4%(具体成绩公布时间未确认);Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Xhigh effort配置,Normalized score 87.1%(具体成绩公布时间未确认);Claude Fable 5.1 · DRACO(100-task Perplexity deep-research benchmark),Adaptive Max配置,Normalized score 87.7%(具体成绩公布时间未确认)。官方出处 Claude Fable 5.1 and Claude Mythos 5.1 System Card(附件,PDF dated September 1, 2026; 212 pages. Full text review completed: pp.1-212. All 64 low-text or figure pages were rendered and visually inspected in 16 sheets; no additional standard Fable 5.1 capability scores were found in the supplementary safety/alignment sections. Revision date unknown. Local: 本地留存副本 and .txt.)本地留存副本, PDF p.179,https://www.anthropic.com/claude-fable-5-1-mythos-5-1-system-card#page=179。Benchlog https://benchlog.lekshome.cc/evidence/claude-fable-and-mythos-5-1/ev-f51-card-draco/
关联成绩与条件
Claude Fable 5.1 · DRACO 成绩与配置
指标与成绩
配置与评测条件
证据关系
Normalized score84.6%原文 84.6% · 单位 %
公布 未确认
Low effort
设置
Claude Fable 5.1; adaptive thinking at low effort.
未知项
Sampling parameters are not specified for each chart point.
工具
允许工具
推理预算
low
框架
未记录
协议
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. low effort.
条件
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1; adaptive thinking at medium effort.
未知项
Sampling parameters are not specified for each chart point.
工具
允许工具
推理预算
medium
框架
未记录
协议
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. medium effort.
条件
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1; adaptive thinking at high effort.
未知项
Sampling parameters are not specified for each chart point.
工具
允许工具
推理预算
high
框架
未记录
协议
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. high effort.
条件
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1;CursorBench 主文引述明确为 max effort;后续 System Card 对 OSWorld/BenchCAD/OfficeQA 对照结果亦标 max effort。
未知项
不同 benchmark 的具体执行细节分别见来源章节;部分测试的采样参数或运行数未单独公布。
工具
允许工具
推理预算
max
框架
未记录
协议
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. max effort.
条件
Single-agent score; no unlabeled points were estimated.
Official source for DRACO · 100-task Perplexity deep-research benchmark in Claude Fable 5.1. See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. Evidence ev-f51-card-draco.
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Low effort, Normalized score 84.6% (exact score publication date unconfirmed). See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. View score conditions
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Medium effort, Normalized score 85.3% (exact score publication date unconfirmed). See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. View score conditions
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration High effort, Normalized score 85.4% (exact score publication date unconfirmed). See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. View score conditions
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Xhigh effort, Normalized score 87.1% (exact score publication date unconfirmed). See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. View score conditions
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Adaptive Max, Normalized score 87.7% (exact score publication date unconfirmed). See “Claude Fable 5.1 and Claude Mythos 5.1 System Card” at 本地留存副本, PDF p.179. View score conditions
Figure 8.12.2.A: Claude Fable 5.1 DRACO normalized scores by effort: low 84.6%, medium 85.3%, high 85.4%, xhigh 87.1%, max 87.7%.
PDF dated September 1, 2026; 212 pages. Full text review completed: pp.1-212. All 64 low-text or figure pages were rendered and visually inspected in 16 sheets; no additional standard Fable 5.1 capability scores were found in the supplementary safety/alignment sections. Revision date unknown. Local: 本地留存副本 and .txt.
Version published
Unconfirmed
Disclosure date
Unconfirmed · Launch or later disclosure unconfirmed
Collected
Historical basis
No reliable historical version evidence available
Citation text
Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Low effort, Normalized score 84.6% (exact score publication date unconfirmed); Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Medium effort, Normalized score 85.3% (exact score publication date unconfirmed); Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration High effort, Normalized score 85.4% (exact score publication date unconfirmed); Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Xhigh effort, Normalized score 87.1% (exact score publication date unconfirmed); Claude Fable 5.1 · DRACO (100-task Perplexity deep-research benchmark), configuration Adaptive Max, Normalized score 87.7% (exact score publication date unconfirmed). Official source: Claude Fable 5.1 and Claude Mythos 5.1 System Card (Attachment, PDF dated September 1, 2026; 212 pages. Full text review completed: pp.1-212. All 64 low-text or figure pages were rendered and visually inspected in 16 sheets; no additional standard Fable 5.1 capability scores were found in the supplementary safety/alignment sections. Revision date unknown. Local: 本地留存副本 and .txt.), 本地留存副本, PDF p.179, https://www.anthropic.com/claude-fable-5-1-mythos-5-1-system-card#page=179. Benchlog https://benchlog.lekshome.cc/evidence/claude-fable-and-mythos-5-1/ev-f51-card-draco/?lang=en
Related scores and conditions
Claude Fable 5.1 · DRACO scores and configurations
Metric and score
Configuration and evaluation conditions
Evidence relation
Normalized score84.6%Reported 84.6% · Unit %
Published Unconfirmed
Low effort
Settings
Claude Fable 5.1; adaptive thinking at low effort.
Unknowns
Sampling parameters are not specified for each chart point.
Tools
Tools allowed
Reasoning budget
low
Framework
not recorded
Protocol
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. low effort.
Conditions
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1; adaptive thinking at medium effort.
Unknowns
Sampling parameters are not specified for each chart point.
Tools
Tools allowed
Reasoning budget
medium
Framework
not recorded
Protocol
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. medium effort.
Conditions
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1; adaptive thinking at high effort.
Unknowns
Sampling parameters are not specified for each chart point.
Tools
Tools allowed
Reasoning budget
high
Framework
not recorded
Protocol
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. high effort.
Conditions
Single-agent score; no unlabeled points were estimated.
Claude Fable 5.1;CursorBench 主文引述明确为 max effort;后续 System Card 对 OSWorld/BenchCAD/OfficeQA 对照结果亦标 max effort。
Unknowns
不同 benchmark 的具体执行细节分别见来源章节;部分测试的采样参数或运行数未单独公布。
Tools
Tools allowed
Reasoning budget
max
Framework
not recorded
Protocol
DRACO 100-task set; web search, web fetch, programmatic tool calling, code execution; 800k-token budget; five independent grading runs per response; normalized score. max effort.
Conditions
Single-agent score; no unlabeled points were estimated.