Llama 3 8B Koalpaca: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1382 ± 17, rank #2035 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 26.26 | Normalized accuracy (%) | 85.5 |
| Open Korean LLM - KorNAT-CKA | 27.6 | Normalized accuracy (%) | 70.7 |
| Open Korean LLM - KorNAT-Harmless | 63.3 | Normalized accuracy (%) | 59.8 |
| Open Korean LLM - Ko-EQ-Bench | 6.23 | EQ-Bench score | 51.8 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 13.8 | Exact match, flexible extract (%) | 41.8 |
| Open Korean LLM Leaderboard | 32.06 | Average Score (%) | 30.8 |
| Open Ko-LLM Leaderboard | 32.06 | Average (%) | 30.6 |
| Open Korean LLM - Ko-IFEval | 24.2 | Strict accuracy, prompt/instruction mean (%) | 29.4 |
| Open Korean LLM - Ko-Winogrande | 52.01 | Accuracy (%) | 8.2 |
| Open Korean LLM - KorNAT-Helpful | 40.4 | Normalized accuracy (%) | 7.4 |
Interactive version: theaggregate.ai/model?slug=llama-3-8b-koalpaca · How It Works · Data refreshed daily, snapshot 2026-09-19.