True Index: Domain Rankings
Head-to-Head: Clean Evals vs. Legacy Benchmarks
| Capability Domain | Clean Evals (Tiers S & A) | Legacy Benchmarks (Tiers C & D) | Public Lift (Boost) |
|---|---|---|---|
| Agentic & Enterprise | 47.6p | 61.3p | +13.7p |
| Coding & SWE | 38.8p | 59.2p | +20.4p |
| General Knowledge & QA | 44.8p | 52.6p | +7.8p |
| Math & Scientific Reasoning | 44.2p | 49.7p | +5.5p |
Audited Benchmark Performance of Qwen3.8 Max (0902)
Grouped by benchmark integrity so private capability signal stands apart from aging public tests.
Top-Tier Validated Tests (Tiers S & A) (13)
Private, leak-resistant tests that build real capability signal.
| Benchmark | Benchmark Integrity Status | Domain | Rank | Percentile |
|---|---|---|---|---|
| AA-Briefcase v1.1 | Private / Fresh | Agentic & Enterprise | #14/50 | 73.5p |
| MLCR-AA | Private / Fresh | Math & Scientific Reasoning | #24/46 | 51.4p |
| FrontierMath | Private / Fresh | Math & Scientific Reasoning | #16/27 | 50.4p |
| MKA (Multimodal Knowledge Agent) | Private / Fresh | Agentic & Enterprise | #16/23 | 46.4p |
| FrontierCode-AA 2.0 | Private / Fresh | Coding & SWE | #28/46 | 41.3p |
| Code2Video-AA | Private / Fresh | Coding & SWE | #14/23 | 40.9p |
| AutomationBench-AA | Private / Fresh | Agentic & Enterprise | #37/50 | 26.5p |
| Code Migration v1.1 | Private / Fresh | Coding & SWE | #10/24 | 60.3p |
| AA-Omniscience v1.1 | Private / Fresh | General Knowledge & QA | #27/47 | 44.8p |
| AA-AnalystAgent | Private / Fresh | Agentic & Enterprise | #16/24 | 44.0p |
| AA-3D | Private / Fresh | Coding & SWE | #19/22 | 33.1p |
| CritPt v2 | Private / Fresh | Math & Scientific Reasoning | #26/40 | 30.7p |
| CyberBench v1.1 | Private / Fresh | Coding & SWE | #25/25 | 18.2p |
Mid-Tier Evaluations (Tier B) (2)
Partially exposed or transitional evaluations with moderate leakage risk.
| Benchmark | Benchmark Integrity Status | Domain | Rank | Percentile |
|---|---|---|---|---|
| HLE-Diamond | Semi-Exposed | General Knowledge & QA | #29/49 | 41.7p |
| VIALS-AA | Semi-Exposed | Math & Scientific Reasoning | #20/23 | 13.6p |
Legacy & Vulnerable Benchmarks (Tiers C & D) (5)
Public or aging benchmarks prone to overfitting and training-data leaks.
| Benchmark | Benchmark Integrity Status | Domain | Rank | Percentile |
|---|---|---|---|---|
| GDPval-AA v2.1 | Public / Aging | Agentic & Enterprise | #14/43 | 65.5p |
| Terminal-Bench 4.0 | Public / Aging | Coding & SWE | #21/50 | 59.2p |
| GDP.pdf | Public / Aging | Agentic & Enterprise | #21/50 | 57.1p |
| Terminal-Bench-Science 0.1 | Public / Aging | Math & Scientific Reasoning | #23/41 | 49.7p |
| MMMU-Pro | Public / Aging | General Knowledge & QA | #10/22 | 52.6p |