BENCHMARK EVIDENCE / ACTIVE
Humanity's Last Exam
multidisciplinary reasoning. Metric: accuracy. Direction: higher-is-better. Results below stay separated by evaluator evidence type and comparable group.
3observations
Artificial Analysisevaluator
activestatus
2026-10-04verified
METHOD
Read the score with its configuration.
Scores from different comparable groups are intentionally not merged into one ranking. Reasoning effort, tools, fallback behavior, evaluator and benchmark version stay attached to every observation.
- Configuration and tool availability must remain attached to observations.
independent
3 observationsartificial-analysis-v4.3.2
| Model | Score | Reasoning | Tools | Source |
|---|---|---|---|---|
| Claude Opus 5.5 | 61.0% | max | benchmark-defined | Evidence ↗ |
| GPT-6 Astra | 53.0% | high | benchmark-defined | Evidence ↗ |
| GPT-6 Sol | 44.0% | high | benchmark-defined | Evidence ↗ |