Qwen 3.6 27B (Reasoning): benchmark results
Provider: Alibaba. Released 2026-04-22. Access: Open.
Aggregate Rating 1634 ± 10, rank #584 of 2168 rated models, from 217 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Meta Muse Glimmer 30B Card - CIMemories - Coverage Rate | 66.9 | Coverage Rate (%) | 100 |
| SEA-HELM (Burmese) - Natural Language Inference | 60.72 | Normalized Score | 100 |
| SEA-HELM (Burmese) - SEA-NLI | 68.1 | Normalized Score | 100 |
| SEA-HELM (Indonesian) - Natural Language Inference | 83.57 | Normalized Score | 100 |
| SEA-HELM (Malay) - SEA-NLI | 71.14 | Normalized Score | 100 |
| SEA-HELM (Tamil) - Causal Reasoning | 95.87 | Normalized Score | 100 |
| SEA-HELM (Tamil) - NLR | 76.81 | Normalized Score | 100 |
| SEA-HELM (Tamil) - Natural Language Inference | 57.76 | Normalized Score | 100 |
| SEA-HELM (Thai) - NLR | 79.45 | Normalized Score | 100 |
| SEA-HELM (Thai) - NLU | 64.67 | Normalized Score | 100 |
| SEA-HELM (Vietnamese) - Causal Reasoning | 95.24 | Normalized Score | 100 |
| SEA-HELM (Vietnamese) - NLR | 79.22 | Normalized Score | 100 |
Interactive version: theaggregate.ai/model?slug=qwen-3-6-27b-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.