List your company

Deep Research on xbench

83AccuracyRE-TRAC-30B-A3B
0.8422.1743.564.83Dec 31, 2025Jan 6, 2026Jan 13, 2026Jan 19, 2026Jan 26, 2026Feb 1, 2026Feb 8, 2026

Evaluation results

MethodLinks
RE-TRAC-30B-A3BModel Category=Intermediate-sized Open-Source Models, Model Size=15B~70B2026.0283—
GPT-5-highModel Category=Closed-Source Models2026.0277.9—
RE-TRAC-4BModel Category=Compact Open-Source Models, Model Size=< 15B2026.0276.6—
Tongyi-DeepResearch-30B-A3BModel Category=Intermediate-sized Open-Source Models, Model Size=15B~70B2026.0275—
NestBrowse-4BModel Category=Compact Open-Source Models, Model Size=< 15B2026.0274—
WebSailor-V2-30B-A3B (RL)Model Category=Intermediate-sized Open-Source Models, Model Size=15B~70B2026.0273.7—
MiniMax-M2-229BModel Category=Large Open-Source Models, Model Size=> 70B2026.0272—
AgentCPM-Explore-4BModel Category=Compact Open-Source Models, Model Size=< 15B2026.0270—
o3Model Category=Closed-Source Models2026.0266.7—
Claude-4.5-SonnetModel Category=Closed-Source Models2026.0266—
WebExplorer-8BModel Category=Compact Open-Source Models, Model Size=< 15B2026.0253.7—
InfoAgent-14BModel Category=Compact Open-Source Models, Model Size=< 15B2026.0240.4—
Youtu-LLM 2BThinking Mode=true, ReAct Framework=true, LLM-as-a-Judge=true2025.1219.519.5
Qwen3 4BThinking Mode=true, ReAct Framework=true, LLM-as-a-Judge=true2025.1218.418.4
Co-SightBase Model=Qwen3-14B2026.0216—
TOOLSELFBase Model=Qwen3-14B2026.0216—
WebSailorBase Model=Qwen3-14B2026.0215—
SmolLM3 3BThinking Mode=true, ReAct Framework=true, LLM-as-a-Judge=true2025.1213.913.9
Vanilla AgentBase Model=Qwen3-14B2026.0213—
Co-SightBase Model=Qwen3-8B2026.0213—
OAgentsBase Model=Qwen3-14B2026.0212—
OWLBase Model=Qwen3-14B2026.0212—
Qwen3 1.7BThinking Mode=true, ReAct Framework=true, LLM-as-a-Judge=true2025.1211.711.7
ReSumBase Model=Qwen3-14B2026.0211—
TOOLSELFBase Model=Qwen3-8B2026.0210—
OAgentsBase Model=Qwen3-8B2026.029—
ReSumBase Model=Qwen3-8B2026.027—
OWLBase Model=Qwen3-8B2026.027—
Vanilla AgentBase Model=Qwen3-8B2026.026—
WebSailorBase Model=Qwen3-8B2026.024—