| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CORD (test) | UDoc | F1 Score98.94 | 136 | 1mo ago | |
| SROIE (test) | StructTexT | F1 Score98.7 | 62 | 4mo ago | |
| FUNSD (test) | LayoutLMv3large | F1 Score92.08 | 55 | 4mo ago | |
| SWDE | Nemotron-Nano 12B v2 | Accuracy92.1 | 28 | 3mo ago | |
| 1,000-note gold-standard set 1.0 (test) | Qwen2.5-14B-Instruct | Micro Precision94.8 | 25 | 2mo ago | |
| FDA | Apriel-1.6 | Accuracy84.5 | 22 | 2mo ago | |
| CORD | UDOP | F1 Score97.6 | 18 | 4mo ago | |
| SROIE | LAMBERT | F1 Score98.17 | 16 | 4mo ago | |
| SROIE v1 (test) | LayoutLM_LARGE | Precision95.24 | 15 | 4mo ago | |
| xbench-deepsearch Out-of-Distribution | Accuracy47 | 14 | 3mo ago | ||
| MeasEval-Ext | MeasHalu-7B | Overall Score57.8 | 13 | 3mo ago | |
| FewN few-shot zero-shot | RETROPROMPT | Accuracy67.3 | 13 | 4mo ago | |
| FUNSD v1 (test) | LayoutLMv3large | F1 Score92.08 | 13 | 4mo ago | |
| TOFU 1% forget | Embedding Space Attack | Cumulative ROUGE-154 | 12 | 4mo ago | |
| ACL Anthology gold benchmark 100 query–chunk pairs | acl-verbatim-modernbert | Word Precision65.43 | 11 | 2mo ago | |
| PolyIE | SCILIRE | Precision17.24 | 10 | 4mo ago | |
| MPEA | Precision (P)50.32 | 10 | 4mo ago | ||
| SciREX | SCILIRE | Precision14.52 | 10 | 4mo ago | |
| TDMS | SCILIRE | Precision18.12 | 10 | 4mo ago | |
| MIT | PrivateLoRA | MIT-D Score95 | 10 | 4mo ago | |
| Real-world documents (test) | Receipts Accuracy94.3 | 10 | 4mo ago | ||
| FiNER (test) | SPO-RS | Accuracy77.5 | 9 | 1mo ago | |
| OpenExp-Action-100 | GPT-5 | Validity Score16.33 | 9 | 3mo ago | |
| Video Subtitles | HunyuanOCR-1.5 | Accuracy93.07 | 8 | 18d ago | |
| Receipts | HunyuanOCR-1.5 | Accuracy92.55 | 8 | 18d ago |