Llama-3-Ko-8B-Instruct: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1417 ± 17, rank #1777 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 30.44 | Normalized accuracy (%) | 83.6 |
| Open Korean LLM - KorNAT-Harmless | 64.91 | Normalized accuracy (%) | 77.6 |
| Open Korean LLM - Ko-IFEval | 33.88 | Strict accuracy, prompt/instruction mean (%) | 69.3 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.74 | Normalized accuracy (%) | 55.6 |
| Open Korean LLM - Ko-EQ-Bench | 7.02 | EQ-Bench score | 53.2 |
| Open Korean LLM - KorNAT-Helpful | 46.9 | Normalized accuracy (%) | 50.3 |
| Open Korean LLM Leaderboard | 34.53 | Average Score (%) | 48.1 |
| Open Ko-LLM Leaderboard | 34.53 | Average (%) | 47.9 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 11.52 | Exact match, flexible extract (%) | 36.8 |
| Open Korean LLM - Ko-Winogrande | 55.96 | Accuracy (%) | 23.7 |
Interactive version: theaggregate.ai/model?slug=llama-3-ko-8b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-19.