Optimization Modeling on ComplexOR (SA)
100Success Approximation (SA)OptiAgent
Evaluation Results
| Method | Links | |
|---|---|---|
| OptiAgentBackbone=Sonnet 4.52026.07 | 100 | |
| OptiAgentBackbone=GPT 5.42026.07 | 100 | |
| GPT 5.4System Prompt=without2026.07 | 83.33 | |
| GPT 5.4System Prompt=with2026.07 | 83.33 | |
| NEMO2026.07 | 77.8 | |
| GPT 5.2System Prompt=with2026.07 | 77.78 | |
| ORLM2026.07 | 77.78 | |
| ORThought2026.07 | 77.78 | |
| LLMOPT2026.07 | 72.7 | |
| DeepSeekSystem Prompt=without2026.07 | 72.22 | |
| GPT 5.2System Prompt=without2026.07 | 66.67 | |
| SonnetSystem Prompt=without2026.07 | 61.11 | |
| Sonnet 4.5System Prompt=with2026.07 | 55.56 | |
| CoE2026.07 | 55.56 | |
| DeepSeek-R1System Prompt=with2026.07 | 50 |