kosolar-hermes-test: benchmark results
Provider: Other. Access: Open.
Unified ELO 1522 ± 17, rank #975 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-EQ-Bench | 38.8 | EQ-Bench score | 89.7 |
| Open Korean LLM Leaderboard | 43.97 | Average Score (%) | 89.7 |
| Open Ko-LLM Leaderboard | 43.97 | Average (%) | 89.6 |
| Open Korean LLM - KorNAT-CKA | 30.74 | Normalized accuracy (%) | 85 |
| Open Korean LLM - Ko-Winogrande | 65.11 | Accuracy (%) | 82.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 42.76 | Exact match, flexible extract (%) | 78.2 |
| Open Korean LLM - Ko-IFEval | 30.09 | Strict accuracy, prompt/instruction mean (%) | 54.4 |
| Open Korean LLM - KorNAT-Harmless | 63.04 | Normalized accuracy (%) | 54.2 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.23 | Normalized accuracy (%) | 48.1 |
| Open Korean LLM - KorNAT-Helpful | 46.08 | Normalized accuracy (%) | 42.1 |
Interactive version: theaggregate.ai/model?slug=kosolar-hermes-test · How It Works · Data refreshed daily, snapshot 2026-09-19.