True Index: Domain Rankings
Agentic & Enterprise92.8p
#2
5 Top-Tier private testsCoding & SWE99.3p
#1
3 Top-Tier private testsMath & Scientific Reasoning97.0p
#1
2 Top-Tier private testsHead-to-Head: Clean Evals vs. Legacy Benchmarks
| Capability Domain | Clean Evals (Tiers S & A) | Legacy Benchmarks (Tiers C & D) | Public Lift (Boost) |
|---|---|---|---|
| Agentic & Enterprise | 92.8p | 87.8p | -5.0p |
| Coding & SWE | 99.3p | 78.0p | -21.3p |
| General Knowledge & QA | 98.4p | 91.4p | -7.0p |
| Math & Scientific Reasoning | 97.0p | 84.8p | -12.2p |
Clean Evals Avg (Tiers S & A)
96.9p
13 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
85.5p
8 public or aging tests
Matched Public Lift (Boost)
-2.5p
Near zero or negative indicates clean generalization
Machine-Readable Export
Audited Benchmark Performance of GPT-6.2 Astra
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 |
|---|---|---|---|---|
| Harvey LAB-AA v1.1 | Private / Fresh | Agentic & Enterprise | #1/21 | 100.0p |
| Vibe Code Bench 1-100 v1.1 | Private / Fresh | Coding & SWE | #1/10 | 100.0p |
| FrontierMath | Private / Fresh | Math & Scientific Reasoning | #1/27 | 99.5p |
| FrontierCode-AA 2.0 | Private / Fresh | Coding & SWE | #3/46 | 97.8p |
| MLCR-AA | Private / Fresh | Math & Scientific Reasoning | #4/46 | 94.4p |
| AutomationBench-AA | Private / Fresh | Agentic & Enterprise | #4/50 | 93.9p |
| MKA (Multimodal Knowledge Agent) | Private / Fresh | Agentic & Enterprise | #5/23 | 91.5p |
| AA-Briefcase v1.1 | Private / Fresh | Agentic & Enterprise | #10/50 | 81.6p |
| AA-3D | Private / Fresh | Coding & SWE | #1/22 | 100.0p |
| ARC-AGI-2 | Private / Fresh | General Knowledge & QA | #1/22 | 100.0p |
| LiveBench | Private / Fresh | General Knowledge & QA | #2/28 | 100.0p |
| AA-AnalystAgent | Private / Fresh | Agentic & Enterprise | #1/24 | 97.2p |
| AA-Omniscience v1.1 | Private / Fresh | General Knowledge & QA | #4/47 | 95.1p |
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 | #2/49 | 97.9p |
| VIALS-AA | Semi-Exposed | Math & Scientific Reasoning | #2/23 | 95.5p |
Legacy & Vulnerable Benchmarks (Tiers C & D) (8)
Public or aging benchmarks prone to overfitting and training-data leaks.
| Benchmark | Benchmark Integrity Status | Domain | Rank | Percentile |
|---|---|---|---|---|
| Autoform-Hard | Public / Aging | Math & Scientific Reasoning | #1/8 | 100.0p |
| Terminal-Bench-Science 0.1 | Public / Aging | Math & Scientific Reasoning | #1/41 | 100.0p |
| BioMysteryBench | Public / Aging | Math & Scientific Reasoning | #3/23 | 96.6p |
| Terminal-Bench 4.0 | Public / Aging | Coding & SWE | #3/50 | 95.9p |
| GDP.pdf | Public / Aging | Agentic & Enterprise | #7/50 | 87.8p |
| SWE Atlas - Codebase QnA | Public / Aging | Coding & SWE | #9/15 | 60.0p |
| HealthBench | Public / Aging | Math & Scientific Reasoning | #14/16 | 42.4p |
| MMMU-Pro | Public / Aging | General Knowledge & QA | #1/22 | 91.4p |