Humanity's Last Exam

Evals

Benchmark / dataset

Frontier expert-level knowledge benchmark.

Category
Autonomy, R&D & self-replication
Insurance exposure
Unmapped
Publisher
Center for AI Safety + Scale AI
Publisher type
Nonprofit / research org
License
MIT
Status
Active (<90d)
Last commit
2026-09-23
Stars
1715
Notes
Capability, not risk; included for context.

View source on GitHub →

Wanich Risk Lab, 2026.