Median Human: benchmark results
General-human benchmark reference assembled from studies with different participants and summary statistics; it does not estimate a population median. Provider: Human.
Aggregate Rating 1956 ± 19, from 44 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BabyVision | 94.1 | Score (%) | 100 |
| Epoch AI - Common Sense Qa 2 | 94.1 | Score | 100 |
| HellaSwag | 95.6 | Accuracy (%) | 100 |
| MERA - MultiQ | 92.8 | F1 (%) | 100 |
| MERA - PARus | 98.2 | Accuracy (%) | 100 |
| MERA - ruHateSpeech | 98.5 | Accuracy (%) | 100 |
| MMSI-Bench | 97.2 | Accuracy (%) | 100 |
| MMToM-QA | 93 | All (%) | 100 |
| MMToM-QA - Belief | 97.5 | Belief (%) | 100 |
| MMToM-QA - Goal | 88.5 | Goal (%) | 100 |
| OpenBookQA | 91.6 | Accuracy (%) | 100 |
| PIQA | 94.9 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=median-human · How It Works · Data refreshed daily, snapshot 2026-10-20.