Asynchronous planning on AsyncHow
98.44Makespan AccuracyDeepSeek-V4-Flash
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-V4-FlashApproach=CP-SAT Formalizer2026.05 | 98.44 | |
| CP-SAT FormalizerSolver=CP-SAT, Averaged=four LLMs2026.05 | 97.5 | |
| GPT-5-miniApproach=CP-SAT Formalizer2026.05 | 97.5 | |
| Qwen3.6 35B A3BApproach=CP-SAT Formalizer2026.05 | 97.19 | |
| Gemini-3-flashApproach=CP-SAT Formalizer2026.05 | 96.88 | |
| GPT-5-miniApproach=Planner2026.05 | 96.56 | |
| DeepSeek-V4-FlashApproach=Planner2026.05 | 96.56 | |
| Gemini-3-flashApproach=PDDL2.1 Formalizer2026.05 | 96.25 | |
| Qwen3.6 35B A3BApproach=Planner2026.05 | 95.63 | |
| PlannerAveraged=four LLMs2026.05 | 94.4 | |
| Gemini-3-flashApproach=Planner2026.05 | 88.75 | |
| GPT-5-miniApproach=PDDL2.1 Formalizer2026.05 | 85 | |
| DeepSeek-V4-FlashApproach=PDDL2.1 Formalizer2026.05 | 80 | |
| PDDL2.1 FormalizerSolver=OPTIC, Averaged=four LLMs2026.05 | 79.3 | |
| Qwen3.6 35B A3BApproach=PDDL2.1 Formalizer2026.05 | 55.94 |