mistral-7B-instruct-ko-test-v0.3: benchmark results
Provider: Mistral. Access: Open.
Unified ELO 1436 ± 17, rank #1624 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 25.76 | Normalized accuracy (%) | 81.3 |
| Open Korean LLM - Ko-IFEval | 37.02 | Strict accuracy, prompt/instruction mean (%) | 77.3 |
| Open Korean LLM - KorNAT-Helpful | 48.85 | Normalized accuracy (%) | 73.6 |
| Open Korean LLM - Ko-EQ-Bench | 24.05 | EQ-Bench score | 69.4 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 24.79 | Exact match, flexible extract (%) | 58.1 |
| Open Ko-LLM Leaderboard | 36.1 | Average (%) | 55.8 |
| Open Korean LLM Leaderboard | 36.1 | Average Score (%) | 55.8 |
| Open Korean LLM - KorNAT-Harmless | 62.9 | Normalized accuracy (%) | 51.5 |
| Open Korean LLM - KorNAT-CKA | 20.02 | Normalized accuracy (%) | 27.7 |
| Open Korean LLM - Ko-Winogrande | 54.38 | Accuracy (%) | 16 |
Interactive version: theaggregate.ai/model?slug=mistral-7b-instruct-ko-test-v0-3 · How It Works · Data refreshed daily, snapshot 2026-09-19.