OmniDocBench 1.5 — leaderboard
Document-understanding benchmark covering OCR, layout parsing, tables, formulas, and information extraction across diverse document types.
Metric: Overall (self-reported). Source: benchmarklist.com. Status: saturated. 50 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.7 Plus | 91.4 |
| 2 | Qwen 3.6 Plus | 91.2 |
| 3 | Gemini 3 Flash (Preview) | 90.37 |
| 4 | Gemini 3.1 Pro (Preview) | 90 |
| 5 | Qwen 3 VL 235B A22B | 89.15 |
| 6 | Gemini 2.5 Pro | 88.03 |
| 7 | Claude Opus 4.6 (Max) | 86.6 |
| 8 | GPT-5.2 | 85.75 |
| 9 | GPT-5.4 | 85.5 |
| 10 | GPT-4o | 75.02 |
| 11 | Gemma 4 E2B | 0.29 |
| 12 | Gemma 4 E4B | 0.18 |
| 13 | Gemma 4 12B | 0.16 |
| 14 | Gemma 4 26B A4B | 0.15 |
| 15 | Gemma 4 31B | 0.13 |
Interactive version: theaggregate.ai/benchmark?slug=omnidocbench-1-5 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.