GDPval
Rank #12 of 51 · 11 models scored · top score 49.7
unknown
Strongest on contamination resistance, weakest on saturation headroom. Caveat: the public test set leaves contamination on the table.
Weighted composite. Each signal is computed from the source-backed leaderboard below.
Does it still separate models?
How far from ceiling / clustered at the top?
Public vs held-out; training-leak risk.
Apples-to-apples, or mixed / vendor-optimized?
How old is the benchmark?
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | GPT-5.2 | 49.7 | unverified | 2025-12-11 |
| 2 | Claude Opus 4.5 | 45.5 | unverified | 2025-11-24 |
| 3 | Claude Opus 4.1 | 43.6 | unverified | 2025-08-05 |
| 4 | Claude Sonnet 4.5 | 42.5 | unverified | 2025-09-29 |
| 5 | Gemini 3 Pro | 40.3 | unverified | 2025-11-18 |
| 6 | GPT-5 | 34.8 | unverified | 2025-08-07 |
| 7 | o3 | 30.8 | unverified | 2024-12-20 |
| 8 | o4-mini | 25.3 | unverified | 2025-04-16 |
| 9 | Gemini 2.5 Pro (Jun 2025) | 23.3 | unverified | 2025-06-05 |
| 10 | Grok 4 | 21.1 | unverified | 2025-07-09 |
| 11 | GPT-4o (Nov 2024) | 9.9 | unverified | 2024-05-13 |