Expert-Level Question Answering on Humanity's Last Exam
40.9AccuracySeed-1.8
Evaluation Results
| Method | Links | |
|---|---|---|
| Seed-1.8Model Access Type=Closed-Source2026.02 | 40.9 | |
| DeepSeek-V3.2Model Access Type=Closed-Source2026.02 | 40.8 | |
| OpenAI-GPT-5-highModel Access Type=Closed-Source2026.02 | 35.2 | |
| Tongyi-DeepResearch 30BModel Access Type=Open-Source2026.02 | 32.9 | |
| Minimax-M2Model Access Type=Closed-Source2026.02 | 31.8 | |
| WebSailor-V2-30B-A3BModel Access Type=Open-Source2026.02 | 30.6 | |
| GLM-4.6Model Access Type=Closed-Source2026.02 | 30.4 | |
| IterResearch-30B-A3BModel Access Type=Open-Source2026.02 | 28.8 | |
| Gemini Deep ResearchModel Access Type=Closed-Source2026.02 | 26.9 | |
| Kimi-ResearcherModel Access Type=Closed-Source2026.02 | 26.9 | |
| Claude-4.5-SonnetModel Access Type=Closed-Source2026.02 | 24.5 | |
| MiroThinker 8BModel Access Type=Open-Source2026.02 | 21.5 | |
| AgentCPM-Explore-4BModel Access Type=Open-Source2026.02 | 19.1 | |
| Merged-Model-4BModel Access Type=Open-Source2026.02 | 17.4 |