Mistral_test03: benchmark results
Provider: Mistral. Access: Open.
Unified ELO 1360 ± 17, rank #2175 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-Helpful | 48.85 | Normalized accuracy (%) | 73.6 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 29.57 | Exact match, flexible extract (%) | 64 |
| Open Korean LLM - Ko-Winogrande | 59.04 | Accuracy (%) | 43.7 |
| Open Korean LLM - KorNAT-Harmless | 62.32 | Normalized accuracy (%) | 41.6 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM Leaderboard | 30.95 | Average Score (%) | 23.2 |
| Open Ko-LLM Leaderboard | 30.95 | Average (%) | 23.1 |
| Open Korean LLM - KorNAT-CKA | 16.69 | Normalized accuracy (%) | 13 |
| Open Korean LLM - Ko-IFEval | 4.72 | Strict accuracy, prompt/instruction mean (%) | 5.9 |
| Open Korean LLM - Ko-GPQA-Diamond | 15.66 | Normalized accuracy (%) | 0 |
Interactive version: theaggregate.ai/model?slug=mistral-test03 · How It Works · Data refreshed daily, snapshot 2026-09-19.