Home Model series Claude Fable Claude Fable · Humanity's Last Exam · no tools;题集版本未注明 · Release history Claude Fable · Humanity's Last Exam · no tools;题集版本未注明 · Release history View scores, configurations, and disclosures for Humanity's Last Exam · no tools;题集版本未注明 across 2 Claude Fable releases, ordered by release date.
Differences are calculated only for unique, active, exact values between direct predecessors when the unit, metric, and explicitly recorded configuration and conditions match.
2026-09-15 · 5 score records for this release, including 5 exact values (may include historical records)
Claude Fable 5.1 · Humanity's Last Exam scores and configurations Metric and score Configuration and evaluation conditions Sources Reported score 60.9% Reported 60.9% no tools · Unit % Published Unconfirmed
Adaptive Max Settings 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 forbidden Reasoning budget max Framework not recorded Protocol Humanity’s Last Exam; no tools; effort max; test-time compute scaling; exact task-set version not stated. Conditions Official comparison table. View this score in rankings Reported score 53.2% Reported 53.2% · 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 forbidden Reasoning budget low Framework not recorded Protocol Humanity’s Last Exam; exact task-set version not stated; tool condition is recorded per row; effort point is from the plotted run. low effort. Conditions No tools; values transcribed from printed chart labels. View this score in rankings Reported score 55.9% Reported 55.9% · Unit % Published Unconfirmed
Medium effort Settings Claude Fable 5.1; adaptive thinking at medium effort. Unknowns Sampling parameters are not specified for each chart point. Tools Tools forbidden Reasoning budget medium Framework not recorded Protocol Humanity’s Last Exam; exact task-set version not stated; tool condition is recorded per row; effort point is from the plotted run. medium effort. Conditions No tools; values transcribed from printed chart labels. View this score in rankings Reported score 58% Reported 58.0% · Unit % Published Unconfirmed
High effort Settings Claude Fable 5.1; adaptive thinking at high effort. Unknowns Sampling parameters are not specified for each chart point. Tools Tools forbidden Reasoning budget high Framework not recorded Protocol Humanity’s Last Exam; exact task-set version not stated; tool condition is recorded per row; effort point is from the plotted run. high effort. Conditions No tools; values transcribed from printed chart labels. View this score in rankings Reported score 60.4% Reported 60.4% · Unit % Published Unconfirmed
Xhigh effort Settings Claude Fable 5.1;Fable 5.2 对照主表将 FrontierCode 2.0 的 32.4% 标为 Xhigh effort。 Unknowns 采样设置与运行数未在主文给出。 Tools Tools forbidden Reasoning budget xhigh Framework not recorded Protocol Humanity’s Last Exam; exact task-set version not stated; tool condition is recorded per row; effort point is from the plotted run. xhigh effort. Conditions No tools; values transcribed from printed chart labels. View this score in rankings
benchmark table > Multidisciplinary reasoning > Humanity's Last Exam Main text · Introducing Claude Mythos and Fable 5.1 本地留存副本, PDF p.178 Attachment · Claude Fable 5.1 and Claude Mythos 5.1 System Card 2026-11-09 · 5 score records for this release, including 5 exact values (may include historical records)
Records for the direct predecessor do not provide matching conditions for a valid difference.
Claude Fable 5.2 · Humanity's Last Exam scores and configurations Metric and score Configuration and evaluation conditions Sources Reported score 65.6% Reported 65.6% · Unit % Published Unconfirmed
HLE no tools / low Settings Claude Fable 5.2;thinking auto;low effort;no tools;980k-token task budget;no context compaction。 Unknowns 该 effort 点的具体 trial 数未在图注列明。 Tools Tools forbidden Reasoning budget low Framework not recorded Protocol HLE without tools; low effort; 980k-token task budget. Conditions Thinking auto; no tools; no context compaction; trial count not stated in chart caption. View this score in rankings Reported score 66.6% Reported 66.6% · Unit % Published Unconfirmed
HLE no tools / medium Settings Claude Fable 5.2;thinking auto;medium effort;no tools;980k-token task budget;no context compaction。 Unknowns 该 effort 点的具体 trial 数未在图注列明。 Tools Tools forbidden Reasoning budget medium Framework not recorded Protocol HLE without tools; medium effort; 980k-token task budget. Conditions Thinking auto; no tools; no context compaction; trial count not stated in chart caption. View this score in rankings Reported score 68.8% Reported 68.8% · Unit % Published Unconfirmed
HLE no tools / high Settings Claude Fable 5.2;thinking auto;high effort;no tools;980k-token task budget;no context compaction。 Unknowns 该 effort 点的具体 trial 数未在图注列明。 Tools Tools forbidden Reasoning budget high Framework not recorded Protocol HLE without tools; high effort; 980k-token task budget. Conditions Thinking auto; no tools; no context compaction; trial count not stated in chart caption. View this score in rankings Reported score 70.7% Reported 70.7% · Unit % Published Unconfirmed
HLE no tools / xhigh Settings Claude Fable 5.2;thinking auto;xhigh effort;no tools;980k-token task budget;no context compaction。 Unknowns 该 effort 点的具体 trial 数未在图注列明。 Tools Tools forbidden Reasoning budget xhigh Framework not recorded Protocol HLE without tools; xhigh effort; 980k-token task budget. Conditions Thinking auto; no tools; no context compaction; trial count not stated in chart caption. View this score in rankings Reported score 73.2% Reported 73.2% · Unit % Published Unconfirmed
HLE no tools / max Settings Claude Fable 5.2;thinking auto;max effort;no tools;980k-token task budget;no context compaction。 Unknowns 该 effort 点的具体 trial 数未在图注列明。 Tools Tools forbidden Reasoning budget max Framework not recorded Protocol HLE without tools; max effort; 980k-token task budget. Conditions Thinking auto; no tools; no context compaction; trial count not stated in chart caption. View this score in rankings
PDF p.218, Figure 8.11.1.B, HLE without tools(已渲染并视觉核读数值标签) Attachment · Claude Fable 5.2 and Claude Mythos 5.2 System Card