DeepSeek-R1-ReDistill-Qwen-7B-v1.1: benchmark results
mobius labs' re-distilled version of DeepSeek-R1-Distill-Qwen-7B, aiming for improved performance over the original distillation, with HQQ quantization support for faster inference. Provider: Mobius Labs. Released 2025-01-27. Access: Open.
Aggregate Rating 1302 ± 25, rank #1479 of 1656 rated models, from 179 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MATH Level 5 | 34.97 | Score | 85.3 |
| Open LLM Leaderboard - MuSR | 40.09 | Score | 42.6 |
| Open Japanese LLM - MR | 74.2 | Score (%) | 38.3 |
| Open Japanese LLM - Jsick Exact Match | 69.96 | Score (%) | 33.4 |
| Open LLM Leaderboard - IFEval | 34.73 | Score | 33.1 |
| Open Arabic LLM - Arabic MMLU Arabic Language (Middle School) | 37.04 | Accuracy (%) | 30.6 |
| Open Japanese LLM - Mbpp Pylint Check | 1.61 | Score (%) | 28.2 |
| Open Arabic LLM - Arabic MMLU HT Machine Learning | 31.25 | Accuracy (%) | 27.2 |
| Open Japanese LLM - Wiki NER SET F1 | 1.77 | Score (%) | 27 |
| Open Japanese LLM - Wiki Coreference SET F1 | 1.14 | Score (%) | 26.3 |
| Open Japanese LLM - Jsem Exact Match | 70.14 | Score (%) | 25.4 |
| Open Japanese LLM - Janli Exact Match | 62.5 | Score (%) | 24.9 |
Interactive version: theaggregate.ai/model?slug=deepseek-r1-redistill-qwen-7b-v1-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.