Qwen 3 30B Instruct: benchmark results
Provider: Alibaba. Released 2025-04-29.
Aggregate Rating 1571 ± 33, rank #1389 of 10916 rated models, from 28 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| PersonalHomeBench - Feature Recommendation (Sole Reasoning) | 26 | MAP@1 (times 100) on 1,000 feature recommendation tasks (ran | 77.8 |
| PersonalHomeBench - Feature Recommendation (Agentic) | 22.9 | MAP@1 (times 100) on 1,000 feature recommendation tasks (ran | 55.6 |
| Med-CoReasoner - Global-MMLU-Medical - German | 77.87 | Accuracy (%) | 50 |
| Med-CoReasoner - MMLU-ProX-Health | 56.42 | Accuracy (%) | 50 |
| Med-CoReasoner - MMLU-ProX-Health - Chinese | 61.86 | Accuracy (%) | 50 |
| Med-CoReasoner - MMLU-ProX-Health - German | 61.72 | Accuracy (%) | 50 |
| Med-CoReasoner - MMLU-ProX-Health - Italian | 63.32 | Accuracy (%) | 50 |
| Med-CoReasoner - MMLU-ProX-Health - Spanish | 64.48 | Accuracy (%) | 50 |
| HOLMES | 48.22 | Overall Accuracy (self-reported) | 40 |
| Med-CoReasoner - Global-MMLU-Medical | 73.86 | Accuracy (%) | 37.5 |
| Med-CoReasoner - Global-MMLU-Medical - Chinese | 77.54 | Accuracy (%) | 37.5 |
| Med-CoReasoner - Global-MMLU-Medical - French | 79.4 | Accuracy (%) | 37.5 |
Interactive version: theaggregate.ai/model?slug=qwen-3-30b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.