Agent Task Performance on τ2-Bench
79.4Retail Performance
SkillComposer-4B
Updated 1mo ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2026.06 | 79.4 | 83.5 | 98.8 | 88 | |
2026.06 | 78.1 | 81.3 | 98.8 | 87 | |
2026.06 | 77.5 | 82.5 | 95.6 | 85.7 | |
2026.06 | 77.5 | 78.8 | 98.8 | 86.3 | |
2026.06 | 77.5 | 80 | 97.5 | 86 | |
2026.06 | 75.6 | 77.5 | 94.4 | 83.5 | |
2026.06 | 74.4 | 78.8 | 98.1 | 84.8 | |
2026.06 | 73.8 | 77.5 | 93.8 | 82.5 | |
2026.06 | 73.8 | 73.8 | 97.5 | 83.3 | |
2026.06 | 72.5 | 77.5 | 93.8 | 82 | |
2026.06 | 72.5 | 76.3 | 92.5 | 81.3 | |
2026.06 | 72.5 | 75 | 92.5 | 81 | |
2026.06 | 72.5 | 73.8 | 96.3 | 82.3 | |
2026.06 | 71.3 | 68.8 | 93.1 | 79.5 | |
2026.06 | 71.3 | 78.8 | 97.5 | 83.3 | |
2026.06 | 70.6 | 80 | 96.3 | 82.8 |