DocScope - Region Grounding: leaderboard
DocScope (2026) evaluates verifiable reasoning over long, visually rich PDF documents: for each human-annotated question of its 730-question test split the model reads the whole document and returns evidence pages, evidence regions, supporting facts and an answer. This board scores the evidence regions on the correctly retrieved pages by strict F1 in percent, with a GPT-5.5 judge deciding whether each gold region is covered by the predicted boxes; higher is better.
Metric: Strict region grounding F1 (%): a GPT-5.5 multimodal judge labels each gold evidence region on the correctly retrieved pages as covered, imprecise or not covered by the predicted boxes, counting only covered over the answerable questions of the 730-question DocScope test split (human-annotated questions on long, visually rich PDF documents; the model receives the whole document and returns evidence pages, evidence regions, supporting facts and an answer), each stage scored only on the gold pages the model retrieved; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 12 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 57.4 | #135 |
| 2 | Claude Sonnet 4.6 | 44.5 | #173 |
| 3 | Claude Opus 4.7 | 42.9 | #85 |
| 4 | Gemini 3.1 Pro (Preview) | 39.7 | #99 |
| 5 | Gemini 3.1 Flash Lite | 36.2 | #371 |
| 6 | Qwen 3.6 Plus | 29.6 | #238 |
| 7 | Qwen 3.5 27B | 28.8 | #375 |
| 8 | Gemma 4 31B | 27.3 | #382 |
| 9 | Qwen 3 VL 235B A22B | 27.2 | #485 |
| 10 | Qwen 3.5 397B A17B | 24.3 | #264 |
| 11 | Qwen 3 VL 32B | 22.7 | #622 |
| 12 | Gemma 4 26B A4B | 7.5 | #590 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=docscope-region-grounding · How It Works · Data refreshed daily, snapshot 2026-11-21.