Home Model series Claude Fable Claude Fable · HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations · Release history Claude Fable · HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations · Release history View scores, configurations, and disclosures for HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations 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 · 2 score records for this release, including 2 exact values (may include historical records)
Claude Fable 5.1 · HealthBench scores and configurations Metric and score Configuration and evaluation conditions Sources Raw score 66.7% Reported 66.7% · 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 HealthBench; average over five trials; adaptive max effort. Conditions Production safety classifiers and refusal fallback to Opus 5; no tools or custom prompts. View this score in rankings Length-adjusted score 60% Reported 60.0% · 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 HealthBench; length-adjusted score; average over five trials; adaptive max effort. Conditions Production safety classifiers and refusal fallback to Opus 5; no tools or custom prompts. View this score in rankings
本地留存副本, PDF p.198 Attachment · Claude Fable 5.1 and Claude Mythos 5.1 System Card 2026-11-09 · 4 score records for this release, including 4 exact values (may include historical records)
Records for the direct predecessor do not provide matching conditions for a valid difference.
Claude Fable 5.2 · HealthBench scores and configurations Metric and score Configuration and evaluation conditions Sources Raw score 58.8% Reported 58.8% · Unit % Published Unconfirmed
Adaptive Max Settings Claude Fable 5.2;adaptive thinking always on;Max effort;默认 temperature/top_p;Card 通用表说明通常五次运行平均,个别测试另行注明。 Unknowns System Card pp.1-264 文字已完整核读;各benchmark章节未明确披露的运行参数仍未知。 Tools Tool access not recorded Reasoning budget max Framework not recorded Protocol HealthBench; Artificial Analysis; GPT-4.1 grader; max effort. Conditions Safety-classifier fallbacks enabled; scores are not comparable with earlier Anthropic system cards graded by Opus 4.8. View this score in rankings Length-adjusted score 59.6% Reported 59.6% · Unit % Published Unconfirmed
Adaptive Max Settings Claude Fable 5.2;adaptive thinking always on;Max effort;默认 temperature/top_p;Card 通用表说明通常五次运行平均,个别测试另行注明。 Unknowns System Card pp.1-264 文字已完整核读;各benchmark章节未明确披露的运行参数仍未知。 Tools Tool access not recorded Reasoning budget max Framework not recorded Protocol HealthBench length-adjusted score; Artificial Analysis; GPT-4.1 grader; max effort. Conditions Safety-classifier fallbacks enabled; scores are not comparable with earlier Anthropic system cards graded by Opus 4.8. View this score in rankings Raw score 60.1% Reported 60.1% · Unit % Published Unconfirmed
Low effort Settings Claude Fable 5.2;adaptive thinking always on;Low effort。 Unknowns none recorded Tools Tool access not recorded Reasoning budget low Framework not recorded Protocol HealthBench; Artificial Analysis; GPT-4.1 grader; low effort. Conditions Safety-classifier fallbacks enabled; scores are not comparable with earlier Anthropic system cards graded by Opus 4.8. View this score in rankings Length-adjusted score 60.4% Reported 60.4% · Unit % Published Unconfirmed
Low effort Settings Claude Fable 5.2;adaptive thinking always on;Low effort。 Unknowns none recorded Tools Tool access not recorded Reasoning budget low Framework not recorded Protocol HealthBench length-adjusted score; Artificial Analysis; GPT-4.1 grader; low effort. Conditions Safety-classifier fallbacks enabled; scores are not comparable with earlier Anthropic system cards graded by Opus 4.8. View this score in rankings
PDF p.245, Figure 8.16.1.A, HealthBench(已渲染并视觉核读数值标签) Attachment · Claude Fable 5.2 and Claude Mythos 5.2 System Card