DrawingVQA - Contextual Interpretation: leaderboard
Metric: Accuracy (%; reasoning depth R2, 34 contextual-interpretation questions of the 92 expert-curated questions on 33 Issued for Construction drawings, multiple choice and open-ended). Source: arxiv.org. Saturation forecast: Around December 2026. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 79.4 |
| 2 | Qwen 3 VL 32B Instruct | 79.4 |
| 3 | Gemini 2.5 Pro | 76.5 |
| 4 | Gemini 2.5 Flash | 73.5 |
| 5 | Claude Haiku 4.5 | 73.5 |
| 6 | Claude Sonnet 4.5 | 64.7 |
| 7 | O3 (2025-04-16) | 61.8 |
| 8 | Qwen 3 VL 8B Instruct | 61.8 |
| 9 | Llama 4 Scout Instruct | 58.8 |
| 10 | InternVL3.5-8B | 58.8 |
| 11 | Qwen 3 VL 30B A3B Instruct | 58.8 |
| 12 | Llama 3.2 11B Instruct | 47.1 |
| 13 | Phi-4 Multimodal Instruct | 47.1 |
| 14 | GPT-4o (2024-08-06) | 44.1 |
Interactive version: theaggregate.ai/benchmark?slug=drawingvqa-contextual-interpretation · How It Works · Data refreshed daily, snapshot 2026-09-29.