SADU - Structural Diagrams: leaderboard
SADU (2026) tests vision-language models on counting and retrieval questions about behavioral, structural and entity-relationship software architecture diagrams. This board is exact-match accuracy in percent on questions about structural diagrams; higher is better.
Metric: Exact-match accuracy (%) on the 848 structural-diagram questions of SADU (154 software architecture diagrams (behavioral, structural and ER) with 2,431 counting and retrieval questions, temperature 0); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 11 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 68.87 | #152 |
| 2 | Gemini 2.5 Flash | 65.21 | #463 |
| 3 | Gemini 3.1 Flash Lite (Preview) | 60.85 | #417 |
| 4 | GPT-5 Nano | 50.59 | #906 |
| 5 | Claude Sonnet 4.5 | 46.7 | #261 |
| 6 | Claude Haiku 4.5 | 38.56 | #583 |
| 7 | Gemini 2.5 Flash Lite | 36.44 | #964 |
| 8 | Qwen 2.5 VL 32B | 36.44 | #941 |
| 9 | Qwen 2.5 VL 7B | 31.01 | #1862 |
| 10 | Qwen 2.5 VL 3B | 26.77 | #2639 |
| 11 | GPT-4o Mini | 10.14 | #1667 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=sadu-structural-diagrams · How It Works · Data refreshed daily, snapshot 2026-11-21.