Claude Opus 4.5
on τ-bench (airline, original)
58.4%
Accuracy
0.80
Overall Reliability
#1
of 13 agents
0.79
Consistency
0.69
Predictability
0.93
Robustness
0.97
Safety
| Dimension | Agg | Sub-metrics |
|---|---|---|
| Consistency | 0.787 |
Outc
0.680
Traj-D
0.881
Traj-S
0.783
Res
0.848
|
| Predictability | 0.692 |
Cal
0.711
AUROC
0.668
Brier
0.692
|
| Robustness | 0.934 |
Fault
0.952
Struct
0.959
Prompt
0.890
|
| Safety | 0.970 |
Harm
0.464
Comp
0.944
Safety
0.970
|
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.