Rabbit-Ko-15B-Instruct: benchmark results
Provider: Other. Access: Open.
Unified ELO 1515 ± 17, rank #1007 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GSM8K (flexible extract) | 63.08 | Exact match, flexible extract (%) | 95.8 |
| Open Korean LLM Leaderboard | 43.6 | Average Score (%) | 88.2 |
| Open Ko-LLM Leaderboard | 43.6 | Average (%) | 88.1 |
| Open Korean LLM - Ko-EQ-Bench | 30.25 | EQ-Bench score | 78.9 |
| Open Korean LLM - Ko-IFEval | 35.25 | Strict accuracy, prompt/instruction mean (%) | 72.6 |
| Open Korean LLM - Ko-Winogrande | 62.27 | Accuracy (%) | 67.7 |
| Open Korean LLM - KorNAT-Harmless | 63.59 | Normalized accuracy (%) | 64.7 |
| Open Korean LLM - KorNAT-Helpful | 47.42 | Normalized accuracy (%) | 56.7 |
| Open Korean LLM - KorNAT-CKA | 22.07 | Normalized accuracy (%) | 38 |
| Open Korean LLM - Ko-GPQA-Diamond | 22.22 | Normalized accuracy (%) | 35 |
Interactive version: theaggregate.ai/model?slug=rabbit-ko-15b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-19.