QwQ 32B-Preview: benchmark results
Alibaba's experimental 32B reasoning preview that pioneered Qwen's long chain-of-thought approach ahead of the full QwQ-32B release (November 2024). Provider: Alibaba. Released 2024-11-27. Access: Open.
Aggregate Rating 1538 ± 11, rank #651 of 1656 rated models, from 274 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open PL LLM - PPC (multiple choice, 5-shot) | 83.2 | Accuracy (%) | 99.7 |
| Open PL LLM - Belebele (generative, 5-shot) | 93.44 | Accuracy (%) | 99.4 |
| Open PL LLM - Belebele (multiple choice, 5-shot) | 92.78 | Accuracy (%) | 99.4 |
| Open PL LLM - PPC (generative, 5-shot) | 82.2 | Accuracy (%) | 99.4 |
| Open LLM Leaderboard - MMLU-Pro | 56.78 | Score | 99 |
| Open PL LLM - Multiple Choice | 65.26 | Average Multiple-Choice Score (%) | 99 |
| Open PL LLM - DYK (generative, 5-shot) | 72.2 | Binary F1 (%) | 98.4 |
| Open PL LLM - KLEJ NER (generative, 5-shot) | 61.22 | Accuracy (%) | 98.4 |
| Open PL LLM - KLEJ NER (multiple choice, 5-shot) | 62.68 | Accuracy (%) | 98.4 |
| Open Japanese LLM - CG | 66.06 | Score (%) | 98.3 |
| Open PL LLM Leaderboard | 67.01 | Average Score (%) | 98.3 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 93.4 | Binary F1 (%) | 97.8 |
Interactive version: theaggregate.ai/model?slug=qwq-32b-preview · How It Works · Data refreshed daily, snapshot 2026-10-20.