Industrial Component Optimization on Industrial Component Optimization (test)
74.5FSRCOSMO-Agent
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| COSMO-AgentScale=8B2026.04 | 74.5 | 87.5 | 76 | 93.5 | 100 | 0.6504 | 6.72 | |
| Gemini-3-FlashScale=–2026.04 | 67.5 | 83 | 75 | 91 | 98 | 0.6802 | 9.32 | |
| Claude-Sonnet-4.5Scale=–2026.04 | 36 | 56 | 70.5 | 74.5 | 92.5 | 0.4809 | 11.25 | |
| Intern-S1Scale=236B2026.04 | 32 | 53 | 75 | 60 | 99.5 | 0.5367 | 7.44 | |
| Qwen3-30BScale=30B2026.04 | 29.5 | 48.5 | 74.5 | 73 | 100 | 0.5789 | 8.6 | |
| Qwen3-NextScale=80B2026.04 | 25.5 | 47.5 | 75 | 58.5 | 99.5 | 0.563 | 8.6 | |
| Llama-4-ScoutScale=17B2026.04 | 21 | 31.5 | 42 | 45.5 | 62.5 | 0.2689 | 2.94 | |
| Intern-S1-miniScale=8B2026.04 | 20 | 24 | 31.5 | 32.5 | 40 | 0.282 | 6.31 |