llama-3-Korean-8B-V3: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1454 ± 17, rank #1501 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-IFEval | 38.11 | Strict accuracy, prompt/instruction mean (%) | 79.6 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 40.11 | Exact match, flexible extract (%) | 74.5 |
| Open Korean LLM - KorNAT-CKA | 27.53 | Normalized accuracy (%) | 70.6 |
| Open Korean LLM Leaderboard | 38.37 | Average Score (%) | 66.7 |
| Open Ko-LLM Leaderboard | 38.37 | Average (%) | 66.6 |
| Open Korean LLM - Ko-EQ-Bench | 6.85 | EQ-Bench score | 52.6 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.23 | Normalized accuracy (%) | 48.1 |
| Open Korean LLM - Ko-Winogrande | 58.09 | Accuracy (%) | 36.3 |
| Open Korean LLM - KorNAT-Helpful | 42.98 | Normalized accuracy (%) | 20.2 |
| Open Korean LLM - KorNAT-Harmless | 60.32 | Normalized accuracy (%) | 13.7 |
Interactive version: theaggregate.ai/model?slug=llama-3-korean-8b-v3 · How It Works · Data refreshed daily, snapshot 2026-09-19.