List your company

Web Navigation on WebWalker Easy

58.75Success RateWebWalker
31.4538.537545.62552.7125Jun 16, 2026

Evaluation results

MethodLinks
WebWalkerLLM Backbone=Qwen-2.5-72B2026.0658.752.7
ReflexionLLM Backbone=Qwen-2.5-72B2026.0657.53.04
ReflexionLLM Backbone=GPT-4o2026.0656.252.91
WebWalkerLLM Backbone=GPT-4o2026.06552.97
WebWalkerLLM Backbone=Qwen-Plus2026.06553.72
StepGuard (+DDPO+CANR)LLM Backbone=Qwen3B2026.0654.253.78
ReActLLM Backbone=GPT-4o2026.0653.752.53
ReflexionLLM Backbone=Qwen-Plus2026.0653.753.66
StepGuard (+DDPO)LLM Backbone=Qwen3B2026.0652.52.85
StepGuard (+CANR)LLM Backbone=Qwen3B2026.0651.813.86
ReActLLM Backbone=Qwen-Plus2026.0648.751.67
ReActLLM Backbone=Qwen-2.5-32B2026.0647.52.21
ReActLLM Backbone=Qwen-2.5-72B2026.0647.51.68
StepGuard (Baseline)LLM Backbone=Qwen3B2026.0647.253.02
ReflexionLLM Backbone=Qwen-2.5-14B2026.0646.252.21
ReflexionLLM Backbone=Qwen-2.5-32B2026.0642.52.52
WebWalkerLLM Backbone=Qwen-2.5-7B2026.0641.253.39
WebWalkerLLM Backbone=Qwen-2.5-14B2026.0641.252.42
WebWalkerLLM Backbone=Qwen-2.5-32B2026.0641.252.69
ReActLLM Backbone=Qwen-2.5-7B2026.0637.53.36
ReflexionLLM Backbone=Qwen-2.5-7B2026.0637.54.03
ReActLLM Backbone=Qwen-2.5-14B2026.0636.251.86
WebWalkerLLM Backbone=Qwen3B2026.0636.253.65
ReflexionLLM Backbone=Qwen3B2026.0633.754.21
ReActLLM Backbone=Qwen3B2026.0632.53.52