Tool Calling on ToolACE 1000 tools
78.72TSAJTPRO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| JTPROBackbone=GPT-52026.04 | 78.72 | 89.26 | 73.55 | |
| JTPROBackbone=GPT-4o mini2026.04 | 75.13 | 83.59 | 63.64 | |
| GEPABackbone=GPT-52026.04 | 75.13 | 86.4 | 67.77 | |
| GEPABackbone=GPT-4o mini2026.04 | 73.39 | 83.36 | 60.33 | |
| JTPROBackbone=o3-mini2026.04 | 71.48 | 87.46 | 64.46 | |
| GEPABackbone=o3-mini2026.04 | 70.11 | 86.59 | 58.68 | |
| BaseBackbone=GPT-52026.04 | 67.658 | 87.35 | 62.366 | |
| BaseBackbone=GPT-4o mini2026.04 | 61.115 | 86.67 | 58.18 | |
| BaseBackbone=o3-mini2026.04 | 58.916 | 85.036 | 51.272 |