Engineering Design on EngDesign Extended
65.4Overall Average ScoreEmbodiedAct
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| EmbodiedActBackbone=GPT-5.22026.02 | 65.4 | 68.6 | 46 | 33.9 | 46.4 | 100 | 46.7 | 82.6 | 86.7 | 66.9 | |
| EmbodiedActBackbone=Claude-Opus-4.52026.02 | 63.8 | 64.1 | 46.3 | 50.7 | 58.2 | 100 | 40 | 86.3 | 83.9 | 59.9 | |
| EmbodiedActBackbone=Gemini-3-Pro2026.02 | 61.2 | 82.6 | 51.7 | 47.5 | 34.7 | 63 | 51.6 | 69.3 | 93.3 | 46.2 | |
| EmbodiedActBackbone=Claude-Sonnet-4.52026.02 | 59.3 | 61.8 | 38.3 | 46.4 | 65.2 | 100 | 40 | 90.1 | 65.6 | 53.7 | |
| EmbodiedActBackbone=GPT-5-mini2026.02 | 58.6 | 75.7 | 28.7 | 47.5 | 36.7 | 63 | 55.2 | 63.9 | 88.9 | 49.2 | |
| CodeActBackbone=Gemini-3-Pro2026.02 | 56.9 | 55.5 | 30.8 | 37.3 | 37 | 84.6 | 40 | 82.9 | 85.6 | 58.4 | |
| EmbodiedActBackbone=Qwen3-235B-Instruct2026.02 | 55.5 | 53.4 | 40.3 | 31.9 | 60.1 | 100 | 26.7 | 68.2 | 61.5 | 60.4 | |
| EmbodiedActBackbone=Gemini-3-Flash2026.02 | 55.4 | 64.2 | 45.3 | 53.8 | 25.5 | 33.3 | 65.7 | 63.6 | 85.6 | 49.4 | |
| Gemini-3-ProParadigm=Generative Models2026.02 | 53.2 | 44.1 | 27 | 35 | 41.3 | 83.5 | 40 | 72.6 | 85.6 | 61.5 | |
| GPT-5.2Paradigm=Generative Models2026.02 | 51.9 | 40.3 | 41.8 | 29.4 | 37 | 52.7 | 40 | 76.9 | 84.1 | 61.6 | |
| CodeActBackbone=Gemini-3-Flash2026.02 | 51.2 | 49.8 | 32.2 | 29 | 34.7 | 73.1 | 46.7 | 67.3 | 96.7 | 45.1 | |
| Gemini-3-FlashParadigm=Generative Models2026.02 | 50.5 | 44.2 | 36.4 | 34.4 | 37 | 64.4 | 40 | 67.9 | 89.4 | 50.7 | |
| CodeActBackbone=Claude-Opus-4.52026.02 | 50.5 | 53.9 | 33.6 | 39 | 33.7 | 74.8 | 40 | 52.8 | 72.6 | 47.9 | |
| CodeActBackbone=GPT-5.22026.02 | 49.4 | 50.3 | 40.7 | 35.5 | 35.1 | 65 | 46.7 | 80.4 | 60.9 | 45.2 | |
| Claude-Opus-4.5Paradigm=Generative Models2026.02 | 48.5 | 41.6 | 31.1 | 51.2 | 31.2 | 64.8 | 40 | 55.3 | 82.8 | 52.3 | |
| GPT-5-miniParadigm=Generative Models2026.02 | 46.9 | 38.3 | 36.3 | 36.4 | 26.7 | 25 | 40 | 82.6 | 85.2 | 41 | |
| CodeActBackbone=Claude-Sonnet-4.52026.02 | 46.5 | 44.4 | 36.1 | 32.7 | 25.1 | 71.5 | 46.7 | 54.9 | 61.1 | 50.6 | |
| CodeActBackbone=GPT-5-mini2026.02 | 45.7 | 45.3 | 31.6 | 29.9 | 27.5 | 77.5 | 60 | 78.6 | 61.9 | 41.1 | |
| Claude-Sonnet-4.5Paradigm=Generative Models2026.02 | 44.5 | 38.2 | 23.4 | 40 | 32.3 | 68.3 | 33.3 | 63.3 | 65.9 | 50.1 | |
| EmbodiedActBackbone=Qwen3-8B-VL-Instruct2026.02 | 44 | 40.9 | 18.6 | 47.5 | 18 | 46.7 | 26.7 | 61.3 | 73.9 | 42.6 | |
| Qwen3-235B-InstructParadigm=Generative Models2026.02 | 42.4 | 33 | 30.5 | 28.9 | 32.6 | 68.8 | 13.3 | 55.2 | 56.7 | 53.1 | |
| CodeActBackbone=Qwen3-235B-Instruct2026.02 | 40.6 | 36 | 33.6 | 26.9 | 29 | 66.7 | 40 | 77.5 | 62.2 | 34.3 | |
| CodeActBackbone=Qwen3-8B-VL-Instruct2026.02 | 38.1 | 28.9 | 35 | 31.5 | 33.8 | 71.2 | 40 | 61.7 | 60.2 | 34 | |
| Qwen3-8B-VL-InstructParadigm=Generative Models2026.02 | 25.9 | 18 | 18.3 | 24.9 | 22.2 | 45.4 | 13.3 | 40.1 | 44.3 | 27.5 |