Qwen2-7B-Multilingual-RP: benchmark results
maywell's multilingual roleplay fine-tune of Qwen2 7B with 32k context, trained on roughly 2B tokens across five languages. Provider: maywell. Released 2024-06-24. Access: Open.
Aggregate Rating 1434 ± 19, rank #1087 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GSM8K (flexible extract) | 54.06 | Exact match, flexible extract (%) | 88 |
| Open Korean LLM Leaderboard | 41.68 | Average Score (%) | 83.2 |
| Open Ko-LLM Leaderboard | 41.68 | Average (%) | 83.1 |
| Open Portuguese LLM - FaQuAD NLI | 77.03 | Macro F1 (%) | 78.4 |
| Open Korean LLM - Ko-IFEval | 37.05 | Strict accuracy, prompt/instruction mean (%) | 77.5 |
| Open LLM Leaderboard - MATH Level 5 | 22.43 | Score | 75.6 |
| Open Portuguese LLM - OAB Exams | 49.75 | Accuracy (%) | 74.6 |
| Open Portuguese LLM - ENEM | 69.21 | Accuracy (%) | 72.6 |
| Open Korean LLM - Ko-EQ-Bench | 24.68 | EQ-Bench score | 70.6 |
| Open LLM Leaderboard - MMLU-Pro | 38.59 | Score | 68.7 |
| Open Portuguese LLM - BLUEX | 57.02 | Accuracy (%) | 68.1 |
| Open Portuguese LLM - ASSIN2 RTE | 91.54 | Macro F1 (%) | 67.9 |
Interactive version: theaggregate.ai/model?slug=qwen2-7b-multilingual-rp · How It Works · Data refreshed daily, snapshot 2026-10-23.