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

Multi-turn Agentic Tool-use on TauBench Retail

50.3Success Rate (SR)

HERO

21.38828.89436.443.906Jun 10, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
50.310.7
2026.06
47.810.8
2026.06
34.79.6
2026.06
34.512.2
2026.06
33.611.9
2026.06
33.313.4
2026.06
30.411.5
2026.06
27.210.5
2026.06
26.112.6
2026.06
22.512.4