MagpieLM-8B-Chat-v0.1: benchmark results
Magpie's MagpieLM-8B-Chat-v0.1, a Llama 3.1 8B aligned via SFT then DPO on Magpie-generated synthetic data, primarily for English use. Provider: Magpie. Released 2024-09-15. Access: Open.
Aggregate Rating 1412 ± 21, rank #1185 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 28.28 | Normalized accuracy (%) | 96.6 |
| Open Korean LLM - KorNAT-Harmless | 67.95 | Normalized accuracy (%) | 91.1 |
| Open Korean LLM - KorNAT-Helpful | 51.55 | Normalized accuracy (%) | 91.1 |
| Open Korean LLM - KorNAT-CKA | 30.11 | Normalized accuracy (%) | 81.9 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 44.2 | Exact match, flexible extract (%) | 79.2 |
| Open Korean LLM Leaderboard | 39.74 | Average Score (%) | 73.8 |
| Open Ko-LLM Leaderboard | 39.74 | Average (%) | 73.5 |
| Open Korean LLM - Ko-Winogrande | 61.25 | Accuracy (%) | 60.9 |
| Open Korean LLM - Ko-IFEval | 30.19 | Strict accuracy, prompt/instruction mean (%) | 54.7 |
| Open Portuguese LLM - BLUEX | 51.18 | Accuracy (%) | 48.8 |
| Open Portuguese LLM - ASSIN2 RTE | 88.81 | Macro F1 (%) | 46.6 |
| Open Portuguese LLM - OAB Exams | 41.23 | Accuracy (%) | 46.2 |
Interactive version: theaggregate.ai/model?slug=magpielm-8b-chat-v0-1 · How It Works · Data refreshed daily, snapshot 2026-10-23.