Tool Use on Simulated 120-tool benchmark 500 tasks 1.0
480Tokens per Turn
B4 CLI Lazy
Updated 1mo ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2026.04 | 480 | 0.94 | 88 | 2.4 | 5.4 | 0.03 | |
2026.04 | 2,368 | 0.91 | 94 | 2 | 4.3 | 0.03 | |
2026.04 | 4,082 | 0.78 | 81 | 2.2 | 4.6 | 0.04 | |
2026.04 | 11,865 | 0.56 | 58 | 3.8 | 7.1 | 0.09 | |
2026.04 | 47,312 | 0.24 | 72 | 4.2 | 7.9 | 0.21 |