List your company

Visual Tool Reasoning on VisualToolBench (test)

12.85Average@4Agent-KB
7.74369.069310.39511.7207Mar 12, 2026

Evaluation results

MethodLinks
Agent-KBModel=Qwen3-VL-235B, Setting=Agent-KB2026.0312.8520.093.88
w/ ToolsModel=Qwen3-VL-235B, Setting=w/ Tools2026.0311.819.623.82
XSkillModel=Qwen3-VL-235B, Setting=XSkill2026.0311.5220.565.12
w/ ToolsModel=Qwen3-VL-32B, Setting=w/ Tools2026.0311.0918.692.65
Dynamic CheatSheetModel=Qwen3-VL-235B, Setting=Dynamic CheatSheet2026.0310.8619.164.09
No ToolsModel=Qwen3-VL-235B, Setting=No Tools2026.0310.5117.76—
Agent-KBModel=Qwen3-VL-32B, Setting=Agent-KB2026.0310.2816.822.88
XSkillModel=Qwen3-VL-32B, Setting=XSkill2026.0310.2819.434.19
Agent Workflow MemoryModel=Qwen3-VL-32B, Setting=Agent Workflow Memory2026.0310.0517.762.73
Agent Workflow MemoryModel=Qwen3-VL-235B, Setting=Agent Workflow Memory2026.0310.0418.223.71
Dynamic CheatSheetModel=Qwen3-VL-32B, Setting=Dynamic CheatSheet2026.039.1115.892.54
No ToolsModel=Qwen3-VL-32B, Setting=No Tools2026.037.9412.62—