Humanity’s Last Exam Benchmark

Explore score history ↗

Humanity’s Last Exam is an AI benchmark in the Benchmark Radar catalog.

Released: 2025-01-27
Openness: unknown
Importer: Benchmark Saturation

Source provenance

Original evidence

Humanity’s Last Exam results and reported scores

Reported benchmark scores · Benchmark Saturation

Reported scores: 5

Model	Organization	Reported value	Reported	Evidence
claude-opus-4-5-20251101-thinking	—	26.32	—	evidence
Gemini 3 pro preview	—	37.72	—	evidence
gpt-5-2025-08-07	—	26.32	—	evidence
gpt-5-pro-2025-10-06	—	33.32	—	evidence
gpt-5.2-2025-12-11	—	28.50	—	evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.

Which sources cite Humanity’s Last Exam?

Humanity’s Last Exam