List your company

Expert-Level Question Answering on Humanity's Last Exam

40.9AccuracySeed-1.8
16.4622.80529.1535.495Feb 6, 2026

Evaluation results

MethodLinks
Seed-1.8Model Access Type=Closed-Source2026.0240.9
DeepSeek-V3.2Model Access Type=Closed-Source2026.0240.8
OpenAI-GPT-5-highModel Access Type=Closed-Source2026.0235.2
Tongyi-DeepResearch 30BModel Access Type=Open-Source2026.0232.9
Minimax-M2Model Access Type=Closed-Source2026.0231.8
WebSailor-V2-30B-A3BModel Access Type=Open-Source2026.0230.6
GLM-4.6Model Access Type=Closed-Source2026.0230.4
IterResearch-30B-A3BModel Access Type=Open-Source2026.0228.8
Gemini Deep ResearchModel Access Type=Closed-Source2026.0226.9
Kimi-ResearcherModel Access Type=Closed-Source2026.0226.9
Claude-4.5-SonnetModel Access Type=Closed-Source2026.0224.5
MiroThinker 8BModel Access Type=Open-Source2026.0221.5
AgentCPM-Explore-4BModel Access Type=Open-Source2026.0219.1
Merged-Model-4BModel Access Type=Open-Source2026.0217.4