General Knowledge and Reasoning on Humanity's Last Exam (HLE) text-only
36.6sHLE ScoreGemini 3 Flash
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini 3 FlashRank=2, Rank change (Δ)=↓ 1, Evaluation Subset=full2026.03 | 36.6 | 16.7 | |
| GPT-5Rank=1, Rank change (Δ)=↑ 1, Evaluation Subset=full2026.03 | 25.3 | 17.6 | |
| DeepSeek-V3.2Rank=4, Rank change (Δ)=↓ 1, Evaluation Subset=text-only subset2026.03 | 25.1 | 12.7 | |
| GPT-5-miniRank=5, Rank change (Δ)=↓ 1, Evaluation Subset=full2026.03 | 19.4 | 11.9 | |
| Grok-4.1-FastRank=7, Rank change (Δ)=↓ 2, Evaluation Subset=full2026.03 | 18.4 | 10.6 | |
| Claude Sonnet 4.5Rank=3, Rank change (Δ)=↑ 3, Evaluation Subset=full2026.03 | 14.7 | 14.2 | |
| Qwen3-30BRank=6, Rank change (Δ)=↑ 1, Evaluation Subset=text-only subset2026.03 | 10.9 | 10.9 | |
| GPT-4.1Rank=8, Rank change (Δ)=→ 0, Evaluation Subset=full2026.03 | 5.2 | 10.6 | |
| GPT-4.1-miniRank=9, Rank change (Δ)=→ 0, Evaluation Subset=full2026.03 | 4.7 | 10.1 | |
| GPT-4.1-nanoRank=10, Rank change (Δ)=→ 0, Evaluation Subset=full2026.03 | 4.5 | 3.6 | |
| Llama-3.1-70BRank=11, Rank change (Δ)=→ 0, Evaluation Subset=text-only subset2026.03 | 3.4 | 0.1 |