Tool-use Agent Scalability and Performance on Tau2-Bench
1RuntimeTau2-Agent
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Tau2-AgentConcurrent tasks=5002025.11 | 1 | 1,519 | 0.5918 | 11,080,385 | 2,654 | |
| P2P-AgentConcurrent tasks=1,5002025.11 | 1 | 22,800 | 0.5921 | 185,376,127 | 41,003 |