Dr. DocBench - Formula CDM: leaderboard

Metric: Character Detection Matching score of display formulas (0-100), 4,514 expert-annotated difficult pages sampled by parser failure from long books in 52 subject domains; each page parsed with the model's own document-to-markdown prompt (specialized parsers through their official APIs); higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 12 models tracked.

Top models

#ModelScore
1GPT-4o36.02
2Qwen 3.5 Flash35.69
3GPT-5.534.55
4Qwen 3.5 122B A10B32.54
5Gemini 3.1 Pro (Preview)32.22
6Claude Opus 4.632.02
7Qwen 3.5 Plus30.4
8Kimi K2.527.04
9Nemotron Nano 12B v2 VL12.82

Interactive version: theaggregate.ai/benchmark?slug=dr-docbench-formula-cdm · How It Works · Data refreshed daily, snapshot 2026-09-29.