llama: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1456 ± 17, rank #1482 of 2656 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-IFEval | 44.12 | Strict accuracy, prompt/instruction mean (%) | 88.5 |
| Open Korean LLM - Ko-EQ-Bench | 33.05 | EQ-Bench score | 82.7 |
| Open Korean LLM - KorNAT-Helpful | 48.52 | Normalized accuracy (%) | 69.6 |
| Open Korean LLM Leaderboard | 38.49 | Average Score (%) | 67.4 |
| Open Ko-LLM Leaderboard | 38.49 | Average (%) | 67.3 |
| Open Korean LLM - KorNAT-CKA | 23.93 | Normalized accuracy (%) | 48.7 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.23 | Normalized accuracy (%) | 48.1 |
| Open Korean LLM - Ko-Winogrande | 59.35 | Accuracy (%) | 45.5 |
| Open Korean LLM - KorNAT-Harmless | 62.09 | Normalized accuracy (%) | 37.8 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 6.07 | Exact match, flexible extract (%) | 26.6 |
| GQA | 0.02 | Overall Accuracy (%) | 17.9 |
Interactive version: theaggregate.ai/model?slug=llama · How It Works · Data refreshed daily, snapshot 2026-09-19.