Calibration

APEX-Agents-AA → Toolathlon

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

Toolathlon is estimated from APEX-Agents-AA with a linear curve fitted on 14 models measured on both: y = 0.5259·x + 0.3295, R² = 0.49, cross-validated error 7.0 pp. It is used for 1 estimate.

Estimated model	APEX-Agents-AA	Toolathlon	Source
Claude Opus 4.6 (Adaptive)	33.0%	50.3%	estimated ± 7.0 pp, low confidence