Llama-3-Open-Ko-8B: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1421 ± 17, rank #1742 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 30.56 | Normalized accuracy (%) | 84.1 |
| Open Korean LLM - Ko-Winogrande | 64.8 | Accuracy (%) | 80.8 |
| Open Korean LLM - Ko-IFEval | 30.2 | Strict accuracy, prompt/instruction mean (%) | 54.9 |
| Open Ko-LLM Leaderboard | 35.22 | Average (%) | 50.5 |
| Open Korean LLM Leaderboard | 35.22 | Average Score (%) | 50.5 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.23 | Normalized accuracy (%) | 48.1 |
| Open Korean LLM - Ko-EQ-Bench | 0.5 | EQ-Bench score | 45.3 |
| Open Korean LLM - KorNAT-Harmless | 62.5 | Normalized accuracy (%) | 44.9 |
| Open Korean LLM - KorNAT-Helpful | 45.2 | Normalized accuracy (%) | 35 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 10.69 | Exact match, flexible extract (%) | 34.8 |
Interactive version: theaggregate.ai/model?slug=llama-3-open-ko-8b · How It Works · Data refreshed daily, snapshot 2026-09-19.