List your company

Multimodal Agent Task on TIR-Bench

47.75Average@4XSkill
18.8926.382533.87541.3675Mar 12, 2026

Evaluation results

MethodLinks
XSkillModel=Gemini-3-Flash2026.0347.7575
Agent-KBModel=Gemini-3-Flash2026.0336.6262
AWMModel=Gemini-3-Flash2026.0334.2562.5
DCModel=Gemini-3-Flash2026.0333.7559
XSkillModel=Gemini-2.5-Pro2026.0333.1258
w/ ToolsModel=Gemini-3-Flash2026.0332.3758.5
XSkillModel=o4-mini2026.0330.2557.5
AWMModel=Gemini-2.5-Pro2026.0329.7553.5
Agent-KBModel=Gemini-2.5-Pro2026.0329.1352.5
No ToolsModel=Gemini-3-Flash2026.0328.554
w/ ToolsModel=Gemini-2.5-Pro2026.0328.3854
XSkillModel=GPT-5-mini2026.0328.2556
DCModel=Gemini-2.5-Pro2026.0327.6251
DCModel=o4-mini2026.0325.552
AWMModel=o4-mini2026.0325.2551
Agent-KBModel=GPT-5-mini2026.0325.1353
Agent-KBModel=o4-mini2026.0324.7552.5
w/ ToolsModel=o4-mini2026.0324.6249.5
DCModel=GPT-5-mini2026.0324.553.5
AWMModel=GPT-5-mini2026.0324.2553
w/ ToolsModel=GPT-5-mini2026.0323.550.5
No ToolsModel=Gemini-2.5-Pro2026.0321.3740
No ToolsModel=o4-mini2026.0320.1345
No ToolsModel=GPT-5-mini2026.032046.5