True Index: Domain Rankings

Private Top-Tier (S & A) standing across 3 core domains
Agentic & Enterprise82.7p
#7
4 Top-Tier private tests
Coding & SWE91.5p
#4
3 Top-Tier private tests
Math & Scientific Reasoning88.2p
#5
1 Top-Tier private test

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 & Enterprise82.7p (4)89.8p (1)+7.1p
Coding & SWE91.5p (3)65.5p (2)-26.0p
Math & Scientific Reasoning88.2p (1)80.0p (4)-8.2p
Clean Evals Avg (Tiers S & A)
83.8p
11 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
78.4p
7 public or aging tests
Matched Public Lift (Boost)
-0.7p
Near zero or negative indicates clean generalization
Machine-Readable Export

Audited Benchmark Performance of GPT-6.2 Sol

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

Top-Tier Validated Tests (Tiers S & A) (11)

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

BenchmarkBenchmark Integrity StatusDomainRankPercentile
Harvey LAB-AA v1.1Private / FreshAgentic & Enterprise#3/2197.7p
FrontierCode-AA 2.0Private / FreshCoding & SWE#5/4693.2p
MLCR-AAPrivate / FreshMath & Scientific Reasoning#7/4688.2p
Vibe Code Bench 1-100 v1.1Private / FreshCoding & SWE#3/1086.4p
AA-Briefcase v1.1Private / FreshAgentic & Enterprise#12/5077.6p
AutomationBench-AAPrivate / FreshAgentic & Enterprise#12/5077.6p
AA-3DPrivate / FreshCoding & SWE#3/2295.0p
LiveBenchPrivate / FreshGeneral Knowledge & QA#9/2881.1p
AA-Omniscience v1.1Private / FreshGeneral Knowledge & QA#11/4779.8p
AA-AnalystAgentPrivate / FreshAgentic & Enterprise#6/2477.7p
ARC-AGI-2Private / FreshGeneral Knowledge & QA#12/2257.5p

Mid-Tier Evaluations (Tier B) (2)

Partially exposed or transitional evaluations with moderate leakage risk.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
HLE-DiamondSemi-ExposedGeneral Knowledge & QA#6/4989.6p
VIALS-AASemi-ExposedMath & Scientific Reasoning#4/2384.1p

Legacy & Vulnerable Benchmarks (Tiers C & D) (7)

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

BenchmarkBenchmark Integrity StatusDomainRankPercentile
Terminal-Bench-Science 0.1Public / AgingMath & Scientific Reasoning#4/4197.5p
Autoform-HardPublic / AgingMath & Scientific Reasoning#2/894.9p
GDP.pdfPublic / AgingAgentic & Enterprise#6/5089.8p
Terminal-Bench 4.0Public / AgingCoding & SWE#6/5089.8p
BioMysteryBenchPublic / AgingMath & Scientific Reasoning#7/2380.6p
HealthBenchPublic / AgingMath & Scientific Reasoning#13/1647.0p
SWE Atlas - Codebase QnAPublic / AgingCoding & SWE#12/1541.2p