Humanity's Last Exam (HLE)
Frontier
2,500 expert-authored questions across math, sciences, and humanities designed to be the hardest closed-ended benchmark for frontier models.
- Publisher
- Center for AI Safety
- Capabilities
- Scientific ReasoningMathFactual Recall
- Format
- Custom
- Size
- 2500 tasks
- License
- MIT
- Published
- Jan 2025
- Updates
- Monthly
- Notable for
- The most cited "what's left to break" benchmark in 2026 — designed as a successor to MMLU when LLMs saturated it; top scores remain under 50% even with frontier reasoning models.
- Canonical
- lastexam.ai
- Official leaderboard
- labs.scale.com/leaderboard/humanitys_last_exam
Cite
Notes
Only stored in your browser.
Top score 61.4% by Claude Opus 5.5 - 476 models reporting (103 frontier)
Score history
474Top models
476
Where it's ranked
1Related tools
4Implementations, trainers, datasets and scaffolds linked to this eval.
Papers
2Contributors
2FAQ
- What is Humanity's Last Exam (HLE)?
- 2,500 expert-authored questions across math, sciences, and humanities designed to be the hardest closed-ended benchmark for frontier models.
- What capabilities does Humanity's Last Exam (HLE) test?
- Humanity's Last Exam (HLE) evaluates scientific reasoning, math, factual recall.
- What is the current top score on Humanity's Last Exam (HLE)?
- The top reported score is 61.4% by Claude Opus 5.5, across 476 models reporting (103 from frontier labs).
- How can a model improve its Humanity's Last Exam (HLE) score?
- Tools linked to Humanity's Last Exam (HLE) on Sophon include HLE, HLE Web Py (Amitprakash2005), HLE Web Py (Prime Community), HLE Web Py (Prime Intellect) - RL environments, datasets, and scaffolds that target this eval.
- What license is Humanity's Last Exam (HLE) under?
- Humanity's Last Exam (HLE) is available under MIT.