Qwen 3.5 27B (Non-reasoning) — benchmark results
Qwen 3.5 27B evaluated with reasoning disabled. Provider: Alibaba. Released 2026-02-24. Access: Open.
Unified ELO 1616 ± 30, rank #442 of 1776 rated models, from 53 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA GPQA Diamond | 84.24 | Accuracy (%) | 81.6 |
| AA TAU-2 Bench | 87.13 | Accuracy (%) | 80.8 |
| AA Terminal-Bench Hard | 31.82 | Accuracy (%) | 75.3 |
| Artificial Analysis Intelligence Index | 29.31 | Intelligence Index | 74.2 |
| UGI - Writing | 39.94 | Writing Score | 73.6 |
| UGI - Natural Intelligence | 28.8 | NatInt Score | 71.7 |
| AA Humanity's Last Exam | 13.21 | Accuracy (%) | 70.9 |
| AA Global-MMLU-Lite - Yoruba | 64.75 | Accuracy (%) | 69.4 |
| AA Global-MMLU-Lite - Indonesian | 87.67 | Accuracy (%) | 66.7 |
| AA Global-MMLU-Lite - Japanese | 87 | Accuracy (%) | 66.5 |
| AA Long Context Reasoning | 55.67 | Accuracy (%) | 65.8 |
| AA Global-MMLU-Lite - German | 88.5 | Accuracy (%) | 64.8 |
Interactive version: theaggregate.ai/model?slug=qwen-3-5-27b-non-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.