DRAGON (Diagrams) - InfographicsVQA: leaderboard
DRAGON (2026) evidence-grounded reasoning on infographics from InfographicsVQA: given the question and its answer, the model must box every diagram region needed to justify the answer. Grounding IoU (%) between predicted and human-verified evidence regions, averaged over three prompting strategies; higher is better.
Metric: Evidence grounding on infographics from InfographicsVQA: given the diagram, the question and the verified answer, the model returns bounding boxes for every visual element needed to justify the answer; Grounding IoU (%) between the union of predicted boxes and the union of the human-verified gold evidence boxes, averaged over the EDGE (direct), SAGE (select then ground) and VERGE (verify and refine) prompting strategies; DRAGON test split; temperature 0.2, top-p 0.7; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 8 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 5.5 | #119 |
| 2 | Claude Opus 4.6 | 4.3 | #113 |
| 3 | Kimi K2.5 | 3.3 | #266 |
| 4 | Claude Sonnet 4.6 | 1.6 | #173 |
| 5 | Llama 4 Maverick Instruct | 1.5 | #1157 |
| 6 | Qwen 3.5 35B A3B | 1.3 | #492 |
| 7 | Gemma 3 27B (IT) | 1.1 | #1324 |
| 8 | InternVL3.5-38B | 1.1 | #1021 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=dragon-diagrams-infographicsvqa · How It Works · Data refreshed daily, snapshot 2026-11-21.