Claude Opus 4.7
on τ-bench (airline, clean)
84.6%
Accuracy
0.89
Overall Reliability
#1
of 15 agents
0.88
Consistency
0.87
Predictability
0.91
Robustness
0.94
Safety
| Dimension | Agg | Sub-metrics |
|---|---|---|
| Consistency | 0.877 |
Outc
0.897
Traj-D
0.938
Traj-S
0.819
Res
0.855
|
| Predictability | 0.873 |
Cal
0.934
AUROC
0.695
Brier
0.873
|
| Robustness | 0.913 |
Fault
0.906
Struct
1.000
Prompt
0.833
|
| Safety | 0.936 |
Harm
0.444
Comp
0.885
Safety
0.936
|
Per-Task Outcome Consistency
Each cell represents a task. Color shows outcome consistency across runs. Hover to see task ID.
Consistency Distribution
KDE of per-task outcome consistency. Peaks at 0 or 1 indicate polarized behavior.
Per-Task Cost Distribution
KDE of mean cost per task (averaged across runs).
Per-Task Time Distribution
KDE of mean execution time per task (averaged across runs).
Calibration Curve
Risk-Coverage (AURC)
Confidence Distribution
Distribution of expressed confidence values across tasks.