Skip to main content
Evaluation sources/Humanity's Last Exam
ObservingKnowledge & language

Humanity's Last Exam

Center for AI Safety / Scale AI · Public model evaluation results from Humanity's Last Exam.

Official evaluation
Source description

This source measures concrete delivery performance in Knowledge & language; raw scores are shown separately from the normalized capability score.

Limitation: Different versions, reasoning settings, and test sets may change scores; read them with the data date.

How it is used

Participating sources affect capability scores; reference, cross-reference, and observing sources are for comparison and verification only.

Model results from this source

Claude Fable 564.564.52026-06-09
Muse Spark 1.162.162.12026-07-09
Claude Fable 55959.02026-06-09
GPT-5.4 Pro58.758.72026-04-23
Claude Opus 4.857.957.92026-06-09
GPT-5.5 Pro57.257.22026-04-23
Claude Opus 4.754.754.72026-04-23
GPT-5.552.252.22026-04-23
GPT-5.452.152.12026-04-23
Claude Opus 4.849.849.82026-06-09
Step 3.7 Flash47.247.22026-05-29
Claude Opus 4.746.946.92026-04-23
Claude Sonnet 4.646.846.82026-06-30
Gemini 3.1 Pro Preview44.444.42026-05-19
GPT-5.5 Pro43.143.12026-04-23
GPT-5.4 Pro42.742.72026-04-23
GPT-5.541.441.42026-04-23
Qwen3.7 Max41.441.42026-05-19
GPT-5.439.839.82026-04-23
Nemotron 3 Ultra 550B A55B37.437.42026-06-04
Claude Sonnet 4.634.634.62026-06-30
Nemotron 3 Ultra 550B A55B26.726.72026-06-04

Frequently asked questions