Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Tool-use on Tau-Bench

67.5TAU-AIR Score

Claude-Sonnet-4

29.0239.014958.99Feb 3, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
67.55447.455.91
2026.02
67.55627.244.99
2026.02
63.874.1296.269.75
2026.02
5877.295.868.69
2026.02
56743450.72
2026.02
5258.7747.1545.29
2026.02
4464.6960.7552.47
2026.02
4163.8266.6751.85
2026.02
4063.3855.9248.99
2026.02
3958.5531.841.5
2026.02
36.558.5541.4541.48
2026.02
33.556.3656.845.92
2026.02
3360.7550.2242.81
2026.02
30.553.2921.9334.74