True Index: Domain Rankings

Private Top-Tier (S & A) standing across 3 core domains
Agentic & Enterprise70.1p
#14
4 Top-Tier private tests
Coding & SWE67.8p
#14
4 Top-Tier private tests
Math & Scientific Reasoning59.9p
#17
3 Top-Tier private tests

Head-to-Head: Clean Evals vs. Legacy Benchmarks

3 matched domains · positive boost means higher legacy scores
Capability Domain Clean Evals (Tiers S & A) Legacy Benchmarks (Tiers C & D) Public Lift (Boost)
Agentic & Enterprise70.1p (4)76.5p (2)+6.5p
Coding & SWE67.8p (4)85.2p (2)+17.4p
Math & Scientific Reasoning59.9p (3)74.9p (2)+15.0p
Non-comparable pre-release tests omitted from audited scoring (model released 2026-09-01): SWE-bench Pro V2 (2026-09-22).
Clean Evals Avg (Tiers S & A)
70.7p
14 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
78.9p
6 public or aging tests
Matched Public Lift (Boost)
+14.2p
Near zero or negative indicates clean generalization
Machine-Readable Export

Audited Benchmark Performance of Claude Fable 5.1

Grouped by benchmark integrity so private capability signal stands apart from aging public tests.

Top-Tier Validated Tests (Tiers S & A) (14)

Private, leak-resistant tests that build real capability signal.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
AA-Briefcase v1.1Private / FreshAgentic & Enterprise#8/5085.7p
FrontierMathPrivate / FreshMath & Scientific Reasoning#9/2775.7p
Code2Video-AAPrivate / FreshCoding & SWE#5/2373.9p
FrontierCode-AA 2.0Private / FreshCoding & SWE#15/4670.7p
Harvey LAB-AA v1.1Private / FreshAgentic & Enterprise#11/2167.4p
AutomationBench-AAPrivate / FreshAgentic & Enterprise#31/5038.8p
MLCR-AAPrivate / FreshMath & Scientific Reasoning#35/4629.4p
LiveBenchPrivate / FreshGeneral Knowledge & QA#5/2894.5p
AA-Omniscience v1.1Private / FreshGeneral Knowledge & QA#6/4789.6p
AA-AnalystAgentPrivate / FreshAgentic & Enterprise#3/2488.3p
AA-3DPrivate / FreshCoding & SWE#8/2275.6p
CritPt v2Private / FreshMath & Scientific Reasoning#8/4074.6p
ARC-AGI-2Private / FreshGeneral Knowledge & QA#9/2271.3p
CyberBench v1.1Private / FreshCoding & SWE#14/2550.9p

Mid-Tier Evaluations (Tier B) (1)

Partially exposed or transitional evaluations with moderate leakage risk.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
HLE-DiamondSemi-ExposedGeneral Knowledge & QA#8/4985.4p

Legacy & Vulnerable Benchmarks (Tiers C & D) (6)

Public or aging benchmarks prone to overfitting and training-data leaks.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
SWE Atlas - Codebase QnAPublic / AgingCoding & SWE#2/15100.0p
GDPval-AA v2.1Public / AgingAgentic & Enterprise#8/4379.6p
HealthBenchPublic / AgingMath & Scientific Reasoning#7/1674.9p
Terminal-Bench-Science 0.1Public / AgingMath & Scientific Reasoning#13/4174.9p
GDP.pdfPublic / AgingAgentic & Enterprise#13/5073.5p
Terminal-Bench 4.0Public / AgingCoding & SWE#15/5070.4p