Synatra-Zephyr-7B-v0.02: benchmark results
Provider: Other. Access: Open.
Unified ELO 1443 ± 17, rank #1562 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-Helpful | 49 | Normalized accuracy (%) | 75.6 |
| Open Korean LLM - Ko-GPQA-Diamond | 24.75 | Normalized accuracy (%) | 69.5 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 27.37 | Exact match, flexible extract (%) | 61.4 |
| Open Korean LLM - Ko-EQ-Bench | 14.69 | EQ-Bench score | 60.4 |
| Open Korean LLM Leaderboard | 37 | Average Score (%) | 60.3 |
| Open Ko-LLM Leaderboard | 37 | Average (%) | 60.2 |
| Open Korean LLM - KorNAT-Harmless | 62.79 | Normalized accuracy (%) | 50.3 |
| Open Korean LLM - Ko-IFEval | 28.53 | Strict accuracy, prompt/instruction mean (%) | 48.4 |
| Open Korean LLM - KorNAT-CKA | 23.82 | Normalized accuracy (%) | 47.3 |
| Open Korean LLM - Ko-Winogrande | 58.33 | Accuracy (%) | 37.8 |
Interactive version: theaggregate.ai/model?slug=synatra-zephyr-7b-v0-02 · How It Works · Data refreshed daily, snapshot 2026-09-19.