List your company

Agentic Tool-use on tau2-bench Retail

82Pass@1CODEDELEGATOR
12.42430.48748.5566.613Jan 21, 2026Feb 10, 2026Mar 2, 2026Mar 22, 2026Apr 11, 2026May 1, 2026May 21, 2026

Evaluation results

MethodLinks
CODEDELEGATORModel=DeepseekV3.22026.018271.263.457———
+PACESLM=Qwen3.5-9B2026.0581.774.569.866.1———
GPT-5Model Category=Proprietary Models2026.0381.671.864.758.8———
CODEDELEGATORModel=GPT5.02026.0180.970.364.560.5———
ReActModel=DeepseekV3.22026.0179.669.963.458.8———
VanillaSLM=Qwen3.5-9B2026.0579.170.665.962.6———
ReActModel=GPT5.02026.0178.768.962.958.8———
+PESLM=Qwen3.5-9B2026.0578.570.465.261.7———
CodeActModel=GPT5.02026.017565.359.357.1———
CodeActModel=DeepseekV3.22026.0170.2595047———
Sonnet 3.52026.0569.257.650.946.2———
GPT-4oModel Category=Proprietary Models2026.036450.84337.7———
Qwen3-235B-A22B-Inst.Model Category=Open-Source Models, Parameter Scale=≥70B2026.0363.651.244.539.5———
Simia-Tau-32BModel Category=Open-Source Models, Parameter Scale=~30B2026.0362.550.141.936———
GPT-4o2026.0560.449.14338.3———
xLAM-2-70b-fc-rModel Category=Open-Source Models, Parameter Scale=≥70B2026.0359.744.135.830.3———
CoVe-4BModel Category=Open-Source Models, Parameter Scale=≤8B2026.0359.4504541.2———
GEM-32BModel Category=Open-Source Models, Parameter Scale=~30B2026.0355.5——————
Qwen3-30B-A3B-Inst.Model Category=Open-Source Models, Parameter Scale=~30B2026.0353.541.234.429.8———
Simia-Tau-8BModel Category=Open-Source Models, Parameter Scale=≤8B2026.0352.439.632.728.1———
Simia-Tau-RL-8BModel Category=Open-Source Models, Parameter Scale=≤8B2026.0352.441.134.930.7———
xLAM-2-32b-fc-rModel Category=Open-Source Models, Parameter Scale=~30B2026.0352.137.129.224.6———
+PACESLM=Qwen3-4B2026.0551.337.830.727———
xLAM-2-8b-fc-rModel Category=Open-Source Models, Parameter Scale=≤8B2026.0348.934.527.522.5———
+PESLM=Qwen3-4B2026.0547.434.127.822.6———
VanillaSLM=Qwen3-4B2026.0546.533.226.121.7———
GEM-8BModel Category=Open-Source Models, Parameter Scale=≤8B2026.0344.5——————
Qwen3-4B-Inst.Model Category=Open-Source Models, Parameter Scale=≤8B2026.0340.127.321.718.4———
Simia-Tau-3BModel Category=Open-Source Models, Parameter Scale=≤8B2026.0332——————
xLAM-2-3b-fc-rModel Category=Open-Source Models, Parameter Scale=≤8B2026.0315.18.86.45.3———
Agent-CBackbone=Qwen3-8B2026.03————42.20234.7
DeepSeek-V3.2Mode=Non-Reasoning, Architecture=MoE, Total Params=671B, Activated Params=37B2026.01————80.9——
EXAONE 4.0Mode=Non-Reasoning, Architecture=Dense, Total Params=32B, Activated Params=32B2026.01————61.4——
K-EXAONEMode=Non-Reasoning, Architecture=MoE, Total Params=236B, Activated Params=23B2026.01————73.2——
LLMBackbone=Qwen3-8B2026.03————11.376.3146.6
Qwen3-235B-A22B Instruct-2507Architecture=MoE, Total Params=235B, Activated Params=22B2026.01————74.6——
SEVerAConstraints=without, Backbone=Qwen3-8B2026.03————49.410.3238.3
SEVerABackbone=Qwen3-8B2026.03————53.60212.4