SADU - Behavioral 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 behavioral diagrams; higher is better.
Metric: Exact-match accuracy (%) on the 810 behavioral-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) | 63.58 | #156 |
| 2 | Gemini 2.5 Flash | 62.1 | #466 |
| 3 | Gemini 3.1 Flash Lite (Preview) | 59.51 | #419 |
| 4 | Claude Sonnet 4.5 | 53.83 | #267 |
| 5 | GPT-5 Nano | 48.64 | #912 |
| 6 | Claude Haiku 4.5 | 48.27 | #587 |
| 7 | Qwen 2.5 VL 32B | 44.44 | #944 |
| 8 | Gemini 2.5 Flash Lite | 42.59 | #972 |
| 9 | Qwen 2.5 VL 7B | 34.2 | #1878 |
| 10 | Qwen 2.5 VL 3B | 29.26 | #2664 |
| 11 | GPT-4o Mini | 19.14 | #1680 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=sadu-behavioral-diagrams · How It Works · Data refreshed daily, snapshot 2026-11-22.