Custom-KoLLM-13B-v3: benchmark results
Provider: Other. Access: Open.
Unified ELO 1384 ± 17, rank #2028 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-Winogrande | 63.3 | Accuracy (%) | 73.4 |
| Open Korean LLM - KorNAT-Helpful | 47.88 | Normalized accuracy (%) | 62.1 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 12.51 | Exact match, flexible extract (%) | 39.2 |
| Open Korean LLM - KorNAT-Harmless | 62.12 | Normalized accuracy (%) | 38.4 |
| Open Korean LLM Leaderboard | 32.4 | Average Score (%) | 33.2 |
| Open Ko-LLM Leaderboard | 32.4 | Average (%) | 33 |
| Open Korean LLM - KorNAT-CKA | 20.74 | Normalized accuracy (%) | 31.1 |
| Open Korean LLM - Ko-IFEval | 24.47 | Strict accuracy, prompt/instruction mean (%) | 30.3 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-GPQA-Diamond | 18.18 | Normalized accuracy (%) | 2.3 |
Interactive version: theaggregate.ai/model?slug=custom-kollm-13b-v3 · How It Works · Data refreshed daily, snapshot 2026-09-19.