List your company

Web Agent Task Completion on Browsecomp

51.5AccuracyOpenAI DeepResearch
0.95614.07827.240.322Jun 5, 2026

Evaluation results

MethodLinks
OpenAI DeepResearchType=Proprietary System2026.0651.5——
OpenAI o3Type=Proprietary System2026.0649.7——
SlimSearcher (SFT+RL)Backbone Policy=Tongyi-DeepResearch, Optimization Strategy=SFT+RL2026.0644.747.6311,093
SlimSearcher (SFT)Backbone Policy=Tongyi-DeepResearch, Optimization Strategy=SFT2026.0642.861.2811,522
Tongyi-DeepResearch2026.064163.712,014
WebLeaperType=Open-weights Agent2026.0638.860.0710,894
Prompt ControlBackbone Policy=Tongyi-DeepResearch, Optimization Strategy=Prompt Control2026.0637.362.812,222
SlimSearcher (SFT+RL)Backbone Policy=Qwen3-30B-A3B-Instruct-2507, Optimization Strategy=SFT+RL2026.0632.847.2210,883
SlimSearcher (SFT)Backbone Policy=Qwen3-30B-A3B-Instruct-2507, Optimization Strategy=SFT2026.0630.965.5312,185
WebExplorerType=Open-weights Agent2026.0613.729.436,289
Claude-4-SonnetType=Proprietary System2026.0612.2——
Kimi-K2-Instruct-0905Type=Open-weights Agent2026.069.416.653,426
Qwen3-235B-A22B-InstructType=Open-weights Agent2026.064.613.71,837
DeepSeek-V3Type=Open-weights Agent2026.063.84.781,317
Qwen3-30B-A3B-Instruct-2507Type=Open-weights Agent, Unified Environment=true2026.062.925.81,683