Dr. DocBench - Formula CDM: leaderboard
Dr. DocBench (2026) parses 4,514 difficult pages sampled by parser failure from long books in 52 subject domains, with expert annotations of text, formulas, tables and reading order. This board is the Character Detection Matching score of display formulas on a 0 to 100 scale; higher is better.
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: Not forecast. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 36.02 | #741 |
| 2 | Qwen 3.5 Flash | 35.69 | #435 |
| 3 | GPT-5.5 | 34.55 | #49 |
| 4 | Qwen 3.5 122B A10B | 32.54 | #331 |
| 5 | Gemini 3.1 Pro (Preview) | 32.22 | #94 |
| 6 | Claude Opus 4.6 | 32.02 | #102 |
| 7 | Qwen 3.5 Plus | 30.4 | #232 |
| 8 | Kimi K2.5 | 27.04 | #256 |
| 9 | Nemotron Nano 12B v2 VL | 12.82 | #2231 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=dr-docbench-formula-cdm · How It Works · Data refreshed daily, snapshot 2026-11-13.