List your company

Long-horizon agentic tasks on BrowseComp-ZH Our Settings

71.3Pass@1AgentSwing
13.37228.41143.4558.489Mar 29, 2026

Evaluation results

MethodLinks
AgentSwingBase Model=DeepSeek-v3.22026.0371.3
DeepSeek-v3.2Context Management=Discard-All2026.0370.2
DeepSeek-v3.2Context Management=Keep-Last-N2026.0369.9
DeepSeek-v3.2Context Management=Summary2026.0369.2
Gemini-3.0-ProContext Management=w/o CM2026.0366.8
MiroThinker-v1.5-30B-A3BContext Management=-2026.0366.8
Claude-4.5-OpusContext Management=w/o CM2026.0362.4
DeepSeek-v3.2Context Management=Baseline (w/o CM)2026.0361.6
OpenAI-o3Context Management=w/o CM2026.0358.1
AgentSwingBase Model=Tongyi-DR-30B-A3B2026.0356.7
Tongyi-DR-30B-A3BContext Management=Discard-All2026.0353.9
Tongyi-DR-30B-A3BContext Management=Keep-Last-N2026.0350.1
Tongyi-DR-30B-A3BContext Management=Summary2026.0349.1
AgentFold-30B-A3BContext Management=-2026.0347.3
Tongyi-DR-30B-A3BContext Management=Baseline (w/o CM)2026.0347.1
IterResearch-30B-A3BContext Management=-2026.0345.2
AgentFounder-30B-A3BContext Management=-2026.0343.3
OpenAI DeepResearchContext Management=-2026.0342.9
DeepMiner-32B-RLContext Management=-2026.0340.1
AgentSwingBase Model=GPT-OSS-120B2026.0338
GPT-OSS-120BContext Management=Keep-Last-N2026.0333.6
GPT-OSS-120BContext Management=Discard-All2026.0331.5
GPT-OSS-120BContext Management=Summary2026.0330.8
GPT-OSS-120BContext Management=Baseline (w/o CM)2026.0328.4
ASearcher-Web-32BContext Management=-2026.0315.6