benchgap
Calibration

CharXiv w/o tools → CharXiv

CharXiv is estimated from CharXiv w/o tools with a Michaelis–Menten curve fitted on 12 models measured on both: y = 2.0000·x / (1.02511 + x), R² = 0.62, cross-validated error 2.5 pp. It is used for 2 estimates.

Estimated modelCharXiv w/o toolsCharXivSource
dots3-note Preview83.1%89.5%estimated ± 2.5 pp, high confidence
Gemini 3.8 Flash86.2%91.4%estimated ± 2.5 pp, high confidence