DocParsingBench - Table: leaderboard
Systems turn 1,400 pages of real business documents (finance, legal, research, manufacturing, education; Chinese and English) into Markdown; the README does not state how many pages each system was scored on. This board is the table part: TEDS tree-edit similarity of each matched HTML table.
Metric: Table TEDS (0-1). Source: github.com. Saturation forecast: Rough model projection: around 2026. 17 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | OXR-1.0 | 0.8004 | |
| 2 | Chandra-OCR-2 | 0.7769 | |
| 3 | TeleOCR | 0.7677 | |
| 4 | MinerU2.5-Pro | 0.7673 | |
| 5 | Dots.mocr | 0.7518 | |
| 6 | Jina-OCR-v1 | 0.7503 | |
| 7 | PaddleOCR-VL-1.6 | 0.7495 | |
| 8 | OvisOCR2 | 0.742 | |
| 9 | Youtu-Parsing | 0.7248 | |
| 10 | Infinity-Parser2-Pro | 0.7248 | |
| 11 | GLM-OCR | 0.7185 | |
| 12 | Qianfan-OCR | 0.6954 | |
| 13 | FireRed-OCR | 0.693 | |
| 14 | Surya OCR 2 | 0.6695 | |
| 15 | LightOnOCR-2-1B | 0.6634 |
Interactive version: theaggregate.ai/benchmark?slug=docparsingbench-table · How It Works · Data refreshed daily, snapshot 2026-11-16.