Calibration
AIME 2025 → AIME26
AIME26 is estimated from AIME 2025 with a linear curve fitted on 6 models measured on both: y = 0.9205·x + 0.0698, R² = 0.99, cross-validated error 4.6 pp. It is used for 5 estimates.
| Estimated model | AIME 2025 | AIME26 | Source |
|---|---|---|---|
| Exaone 4.0 32B | 85.3% | 85.5% | estimated ± 4.6 pp, medium confidence |
| Kimi K2.5 (Reasoning) | 96.1% | 95.4% | estimated ± 4.6 pp, medium confidence |
| LFM2.5-2.6B | 51.9% | 54.7% | estimated ± 4.6 pp, medium confidence |
| MiMo-V2-Flash | 94.1% | 93.6% | estimated ± 4.6 pp, medium confidence |
| Nemotron 3 Nano Omni 30B A3B | 82.1% | 82.6% | estimated ± 4.6 pp, medium confidence |