polyglot-ko-3.8B: benchmark results
Provider: Other. Access: Open.
Unified ELO 1365 ± 18, rank #2138 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 26.63 | Normalized accuracy (%) | 65.8 |
| Open Korean LLM - Ko-IFEval | 33.17 | Strict accuracy, prompt/instruction mean (%) | 65.1 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.74 | Normalized accuracy (%) | 55.6 |
| Open Korean LLM Leaderboard | 31.43 | Average Score (%) | 27.4 |
| Open Ko-LLM Leaderboard | 31.43 | Average (%) | 27.3 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-Winogrande | 55.49 | Accuracy (%) | 20.7 |
| Open Korean LLM - KorNAT-Harmless | 59.78 | Normalized accuracy (%) | 9.1 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 0.3 | Exact match, flexible extract (%) | 8 |
| Open Korean LLM - KorNAT-Helpful | 39.3 | Normalized accuracy (%) | 5 |
Interactive version: theaggregate.ai/model?slug=polyglot-ko-3-8b · How It Works · Data refreshed daily, snapshot 2026-09-19.