Long-horizon planning on TravelPlanner
100Delivery RateAgentVerse
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| AgentVerseMode=Two-Stage2026.02 | 100 | 66.7 | 4.8 | 15.2 | 10.6 | 2.2 | |
| CoTMode=Sole-Planning2026.02 | 100 | 75.8 | 13.3 | 55.2 | 48.9 | 6.7 | |
| AgentVerseMode=Sole-Planning2026.02 | 100 | 79.2 | 11.1 | 59 | 53.3 | 6.7 | |
| SwarmAgenticMode=Sole-Planning2026.02 | 100 | 92.9 | 56.1 | 66.7 | 52.8 | 32.2 | |
| MASFlyMode=Sole-Planning2026.02 | 98.9 | 92.9 | 68.3 | 68.3 | 65.6 | 61.7 | |
| AgentSquareMode=Two-Stage2026.02 | 98.3 | 81.2 | 33.3 | 20.5 | 15.6 | 13.3 | |
| MetaGPTMode=Sole-Planning2026.02 | 98.3 | 82.5 | 22.2 | 60.5 | 50.5 | 13.3 | |
| MetaGPTMode=Two-Stage2026.02 | 95.6 | 65.3 | 14.4 | 19.8 | 16.1 | 8.9 | |
| MASFlyMode=Two-Stage2026.02 | 95.6 | 85.3 | 47.8 | 49.5 | 44.4 | 39.4 | |
| EvoAgentMode=Sole-Planning2026.02 | 95.6 | 69.2 | 13.9 | 59.5 | 47.8 | 8.3 | |
| MegaAgentMode=Sole-Planning2026.02 | 94.4 | 82.9 | 26.2 | 60.5 | 46.9 | 10 | |
| ReActMode=Two-Stage2026.02 | 90.6 | 60.4 | 7.2 | 24 | 16.6 | 5.6 | |
| ReflexionMode=Two-Stage2026.02 | 89.4 | 62.8 | 7.2 | 26.5 | 17.2 | 8.9 |