LLM Stats (Humanity's Last Exam (no tools, text-only)): leaderboard
LLM Stats mirror of Humanity's Last Exam restricted to text questions, evaluated without tools.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around July 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 64.4 | #3 |
| 2 | Claude Sonnet 5.5 | 56.9 | #14 |
| 3 | Hy3 | 47 | #100 |
| 4 | Claude Haiku 5.5 | 45.9 | #43 |
| 5 | DeepSeek V4.1 Flash (1017) | 44 | #27 |
| 6 | MiniMax-M3.1-Flash | 43.4 | #52 |
| 7 | Hy4 preview | 43.4 | #45 |
| 8 | DeepSeek V4 Pro (0813) | 42.7 | #58 |
| 9 | DeepSeek V4.1 Flash | 39.1 | #37 |
| 10 | Qwen 3 Max (Thinking) | 30.2 | #213 (Qwen 3 Max) |
| 11 | Inkling | 29.7 | #144 |
| 12 | Seed 2.0 Mini | 13.3 | #279 |
No result here: #6 GPT-6 Astra, #7 GPT-6.1 Sol, #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-humanity-s-last-exam-no-tools-text-only · How It Works · Data refreshed daily, snapshot 2026-11-05.