List your company
ResearchFormerly WizWand

Autonomous Agent Problem Solving on Humanity's Last Exam

41.6Avg@3ChatGPT-Agent
16.32822.88929.4536.011Nov 14, 2025
Updated 6mo ago

Evaluation results

MethodLinks
ChatGPT-AgentType=Research Agents2025.1141.6
MiroThinker-v1.0-72BParameters=72B2025.1137.7
OpenAI-GPT-5-highType=Foundation Models with Tools2025.1135.2
MiroThinker-v1.0-30BParameters=30B2025.1133.4
Tongyi-DeepResearch-30BType=Research Agents2025.1132.9
Minimax-M2Type=Foundation Models with Tools2025.1131.8
GLM-4.6Type=Foundation Models with Tools2025.1130.4
DeepSeek-V3.1Type=Foundation Models with Tools2025.1129.8
SFR-DeepResearch-20BType=Research Agents2025.1128.7
DeepSeek-V3.2Type=Foundation Models with Tools2025.1127.2
Kimi-ResearcherType=Research Agents2025.1126.9
OpenAI DeepResearchType=Research Agents2025.1126.6
OpenAI-o3Type=Foundation Models with Tools2025.1124.9
Claude-4.5-SonnetType=Foundation Models with Tools2025.1124.5
Kimi-K2-0905Type=Foundation Models with Tools2025.1121.7
MiroThinker-v1.0-8BParameters=8B2025.1121.5
Claude-4-SonnetType=Foundation Models with Tools2025.1120.3
AFM-32B-RLType=Research Agents2025.1118
WebExplorer-8B-RLType=Research Agents2025.1117.3