Mathematical Reasoning on Humanity's Last Exam Math (test)
29.76AccuracySkill-MAS-optimized
Evaluation Results
| Method | Links | |
|---|---|---|
| Skill-MAS-optimizedMeta-agent=Qwen3.5-Plus2026.06 | 29.76 | |
| EvoAgentMeta-agent=Qwen3.5-Plus2026.06 | 27.38 | |
| Skill-MAS-optimizedMeta-agent=DeepSeek-V4-Flash2026.06 | 26.79 | |
| AFlowMeta-agent=Qwen3.5-Plus2026.06 | 25 | |
| Skill-MAS-initMeta-agent=Qwen3.5-Plus2026.06 | 24.4 | |
| Skill-MAS-optimizedMeta-agent=Gemini-3.1-Flash2026.06 | 21.43 | |
| MAS-OrchestraMeta-agent=Qwen3.5-Plus2026.06 | 21.43 | |
| AFlowMeta-agent=DeepSeek-V4-Flash2026.06 | 21.43 | |
| Skill-MAS-initMeta-agent=DeepSeek-V4-Flash2026.06 | 20.83 | |
| MAS-OrchestraMeta-agent=DeepSeek-V4-Flash2026.06 | 20.24 | |
| AOrchestraMeta-agent=DeepSeek-V4-Flash2026.06 | 19.64 | |
| Skill-MAS-optimizedMeta-agent=GPT-5.4-Nano2026.06 | 18.45 | |
| MAS-OrchestraMeta-agent=Gemini-3.1-Flash2026.06 | 17.26 | |
| MAS2Meta-agent=Qwen3.5-Plus2026.06 | 16.56 | |
| Skill-MAS-initMeta-agent=Gemini-3.1-Flash2026.06 | 14.88 | |
| MAS-OrchestraMeta-agent=GPT-5.4-Nano2026.06 | 14.88 | |
| EvoAgentMeta-agent=GPT-5.4-Nano2026.06 | 14.29 | |
| Skill-MAS-initMeta-agent=GPT-5.4-Nano2026.06 | 12.5 | |
| AOrchestraMeta-agent=Qwen3.5-Plus2026.06 | 12.5 | |
| AFlowMeta-agent=GPT-5.4-Nano2026.06 | 11.31 | |
| MAS2Meta-agent=GPT-5.4-Nano2026.06 | 10.42 | |
| EvoAgentMeta-agent=DeepSeek-V4-Flash2026.06 | 9.52 | |
| EvoAgentMeta-agent=Gemini-3.1-Flash2026.06 | 8.93 | |
| AOrchestraMeta-agent=GPT-5.4-Nano2026.06 | 8.93 | |
| AOrchestraMeta-agent=Gemini-3.1-Flash2026.06 | 7.14 | |
| MAS2Meta-agent=DeepSeek-V4-Flash2026.06 | 6.67 | |
| AFlowMeta-agent=Gemini-3.1-Flash2026.06 | 5.95 | |
| MAS2Meta-agent=Gemini-3.1-Flash2026.06 | 2.33 |