Qwen 1.5 32B: benchmark results
Alibaba's 32B mid-size addition to the open-weights Qwen1.5 series (April 2024), adding GQA and 32K context below the 72B tier. Provider: Alibaba. Released 2024-04-02. Access: Open.
Aggregate Rating 1497 ± 12, rank #794 of 1656 rated models, from 337 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Lite - MATH Level 1 - Intermediate Algebra | 84.62 | Equivalent (CoT) (%) | 98.3 |
| Open PL LLM - KLEJ NER (multiple choice, 5-shot) | 60.54 | Accuracy (%) | 96.6 |
| Open Japanese LLM - Xlsum JA RougeLsum | 29.87 | Score (%) | 95.7 |
| Open Japanese LLM - Xlsum JA Bert Score JA F1 | 71.47 | Score (%) | 95 |
| Open Japanese LLM - Xlsum JA Rouge1 | 35.64 | Score (%) | 94.7 |
| Open Japanese LLM - SUM | 12.61 | Score (%) | 94 |
| Open Japanese LLM - Xlsum JA Rouge2 | 12.61 | Score (%) | 94 |
| HELM NaturalQuestions (Open) | 77.71 | F1 (%) | 90 |
| Open Chinese LLM - C-Eval Semantic | 88.88 | Accuracy (%) | 89.3 |
| Open Chinese LLM - GSM8K | 64.44 | Accuracy (%) | 88.8 |
| Open Chinese LLM - WinoGrande | 68.98 | Accuracy (%) | 88.8 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 87.95 | Binary F1 (%) | 88.2 |
Interactive version: theaggregate.ai/model?slug=qwen-1-5-32b · How It Works · Data refreshed daily, snapshot 2026-10-20.