mistralopithecus-v1-dpo: benchmark results
Provider: Mistral. Access: Open.
Unified ELO 1477 ± 17, rank #1318 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 34.79 | Normalized accuracy (%) | 94.6 |
| Open Korean LLM - KorNAT-Helpful | 52.45 | Normalized accuracy (%) | 93 |
| Open Korean LLM - Ko-GPQA-Diamond | 27.27 | Normalized accuracy (%) | 92.4 |
| Open Korean LLM Leaderboard | 39.64 | Average Score (%) | 73.1 |
| Open Ko-LLM Leaderboard | 39.64 | Average (%) | 72.8 |
| Open Korean LLM - Ko-Winogrande | 63.06 | Accuracy (%) | 72.1 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 31.84 | Exact match, flexible extract (%) | 65.8 |
| Open Korean LLM - Ko-EQ-Bench | 17.16 | EQ-Bench score | 63 |
| Open Korean LLM - Ko-IFEval | 30.25 | Strict accuracy, prompt/instruction mean (%) | 55.3 |
| Open Korean LLM - KorNAT-Harmless | 62.43 | Normalized accuracy (%) | 43.9 |
Interactive version: theaggregate.ai/model?slug=mistralopithecus-v1-dpo · How It Works · Data refreshed daily, snapshot 2026-09-19.