Optimization modeling and solving on OptMATH-Bench
61.45Pass@1 (SA)OptSkills-D
Evaluation Results
| Method | Links | |
|---|---|---|
| OptSkills-DCategory=Skill-Based Methods, Base Model=DeepSeek-V3.22026.05 | 61.45 | |
| Gemini-3.1-ProCategory=General Models2026.05 | 54.22 | |
| OptSkills-QCategory=Skill-Based Methods, Base Model=Qwen3-235B-A22b-instruct-25072026.05 | 54.22 | |
| Trace2SkillCategory=Skill-Based Methods2026.05 | 49.4 | |
| GPT-5.4Category=General Models2026.05 | 46.39 | |
| DeepSeek-V3.2Category=General Models2026.05 | 40.36 | |
| Qwen3-235BCategory=General Models2026.05 | 39.76 | |
| AlphaOPTCategory=Experience-Enhanced Methods2026.05 | 36.75 | |
| Chain-of-ExpertsCategory=Agent-Based Methods2026.05 | 35.54 | |
| ORThoughtCategory=Agent-Based Methods2026.05 | 26.51 | |
| OptiMUSCategory=Agent-Based Methods2026.05 | 22.89 | |
| LEAN-LLM-OPTCategory=Experience-Enhanced Methods2026.05 | 22.89 | |
| ORMindCategory=Agent-Based Methods2026.05 | 21.69 |