Calibration

Artificial Analysis Intelligence Index → AA-HLE

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

AA-HLE is estimated from Artificial Analysis Intelligence Index with a Hill curve fitted on 195 models measured on both: y = 0.0336 + (0.5913 − 0.0336)·x^2.48 / (0.27433^2.48 + x^2.48), R² = 0.91, cross-validated error 4.4 pp. It is used for 2 estimates.

Estimated model	Artificial Analysis Intelligence Index	AA-HLE	Source
Claude 4.1 Opus	18.6%	18.7%	estimated ± 4.4 pp, high confidence
o1-pro	12.4%	10.2%	estimated ± 4.4 pp, high confidence