Flowchart to Mermaid code translation on FLOWVQA (test)
98.6Entity PrecisionGPT-4.1-mini
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| GPT-4.1-mini2025.12 | 98.6 | 98.6 | 98.6 | 98 | 97.8 | 97.9 | 0.858 | |
| Gemini-2.5-Flash2025.12 | 98.6 | 98.7 | 98.6 | 97.6 | 97.6 | 97.6 | 0.863 | |
| GPT-4.12025.12 | 97.5 | 97.4 | 97.5 | 96.3 | 96 | 96.1 | 0.875 | |
| GPT-4o2025.12 | 94.9 | 94.6 | 94.6 | 91.9 | 90.9 | 91.4 | 0.901 | |
| GPT-4o-mini2025.12 | 82.7 | 80.8 | 81.7 | 75.9 | 72.9 | 74.3 | 0.908 |