QwQ-32B-Coder-Fusion-9010: benchmark results
huihui-ai's weight blend of QwQ-32B-Preview-abliterated (90%) and Qwen2.5-Coder-32B-Instruct-abliterated (10%). It remains distinct from both parents and the separately published 80/20 and 70/30 blends. Provider: huihui-ai. Access: Open.
Aggregate Rating 1573 ± 33, rank #1364 of 10916 rated models, from 61 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - CG | 67.87 | Score (%) | 99.3 |
| Open LLM Leaderboard - MMLU-Pro | 56 | Score | 98.8 |
| Open LLM Leaderboard - MATH Level 5 | 53.17 | Score | 98.3 |
| Open Portuguese LLM - BLUEX | 77.61 | Accuracy (%) | 96.8 |
| Open Portuguese LLM - ENEM | 83.14 | Accuracy (%) | 96.3 |
| Open LLM Leaderboard - BBH | 67.27 | Score | 95.7 |
| Open Japanese LLM - Jmmlu Exact Match | 76.16 | Score (%) | 95.6 |
| Open Portuguese LLM - OAB Exams | 62.78 | Accuracy (%) | 95.2 |
| Open Japanese LLM Leaderboard | 66.87 | Average Score (%) | 94 |
| Open Japanese LLM - HE | 78 | Score (%) | 93.5 |
| Open Japanese LLM - Wiki NER SET F1 | 14.16 | Score (%) | 93.1 |
| Open Japanese LLM - Xlsum JA Bert Score JA F1 | 71.07 | Score (%) | 92.2 |
Interactive version: theaggregate.ai/model?slug=qwq-32b-coder-fusion-9010 · How It Works · Data refreshed daily, snapshot 2026-11-08.