ResearchBenchmarksAutonomous Agent Task Solving on AgentBoard Tool-query WeatherFollow75Task CompletenessAdMem2336.55063.5Jun 5, 2026Evaluation ResultsMethodMethodLinksTask CompletenessAverage ProgressAdMemLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.067590.36ReActLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.066085.65AWMLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.062575.29