Calibration

We-Math → OmniDocBench 1.5

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap
License: https://creativecommons.org/licenses/by-nc/4.0/

OmniDocBench 1.5 is estimated from We-Math with a offset logistic curve fitted on 5 models measured on both: y = 0.8687 + (1.2000 − 0.8687) / (1 + exp(−42.14·(x − 0.9290))), R² = 0.71, cross-validated error 2.8 pp. It is used for 0 estimates.

Fitted on

Model	We-Math	OmniDocBench 1.5
Claude Opus 4.5	70.0%	87.7%
Gemini 3 Pro	86.9%	88.5%
GPT-5.2	79.0%	85.7%
Kimi K2.5	84.7%	88.8%
Qwen3.5 397B	87.9%	90.8%