Calibration

Artificial Analysis Intelligence Index → Vals GPQA Diamond

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

Vals GPQA Diamond is estimated from Artificial Analysis Intelligence Index with a Hill curve fitted on 47 models measured on both: y = 0.0000 + (0.9158 − 0.0000)·x^5.17 / (0.13558^5.17 + x^5.17), R² = 0.76, cross-validated error 7.5 pp. It is used for 5 estimates.

Estimated model	Artificial Analysis Intelligence Index	Vals GPQA Diamond	Source
Claude 4.1 Opus	18.6%	76.5%	estimated ± 7.5 pp, medium confidence
GPT-4 Turbo	7.0%	3.0%	estimated ± 7.5 pp, low confidence
MiMo-V2.6-Flash	37.9%	91.1%	estimated ± 7.5 pp, medium confidence
MiMo-V2.6-Pro	46.3%	91.4%	estimated ± 7.5 pp, medium confidence
Mistral Large 4	38.4%	91.2%	estimated ± 7.5 pp, medium confidence