mt0-xxl: benchmark results
Provider: Other. Access: Open.
Unified ELO 1308 ± 17, rank #2430 of 2656 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Greek MMLU | 56.57 | Accuracy (%) | 48.1 |
| Greek MMLU - Social Sciences | 61.33 | Accuracy (%) | 48.1 |
| Greek MMLU - STEM | 52.72 | Accuracy (%) | 46.9 |
| Greek MMLU - Greek-Specific | 56.91 | Accuracy (%) | 45.7 |
| Greek MMLU - Humanities | 53.13 | Accuracy (%) | 40.7 |
| Open Korean LLM - Ko-GPQA-Diamond | 22.22 | Normalized accuracy (%) | 35 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-IFEval | 22.59 | Strict accuracy, prompt/instruction mean (%) | 23.9 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 4.25 | Exact match, flexible extract (%) | 22.7 |
| Greek MMLU - Other | 36.92 | Accuracy (%) | 22.2 |
| Open Korean LLM - Ko-Winogrande | 53.43 | Accuracy (%) | 12.6 |
| Open Korean LLM Leaderboard | 28.61 | Average Score (%) | 7.2 |
Interactive version: theaggregate.ai/model?slug=mt0-xxl · How It Works · Data refreshed daily, snapshot 2026-09-19.