Qwen 3.6 27B: benchmark results
Alibaba's dense open Qwen 3.6 27B model, tuned for agentic coding. Provider: Alibaba. Released 2026-04-22. Access: Open.
Aggregate Rating 1711 ± 4, rank #170 of 1656 rated models, from 1236 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CAT — MMLU zero-shot — Base | 84.5 | Accuracy (%) | 100 |
| CAT — MMLU zero-shot — StTP | 83.9 | Accuracy (%) | 100 |
| CAT — MT-Bench — Base | 8.59 | Judge score (1–10) | 100 |
| CAT — MT-Bench — Honest persona | 7.85 | Judge score (1–10) | 100 |
| CAT — MT-Bench — Honest persona + prefill | 8.43 | Judge score (1–10) | 100 |
| CAT — MT-Bench — Honest system | 9.06 | Judge score (1–10) | 100 |
| CAT — MT-Bench — StTP | 8.3 | Judge score (1–10) | 100 |
| CATTest - Pass/Not-Pass Precision | 89.06 | Precision (%) | 100 |
| CheXpercept | 92.2 | Stage 1 (End-to-End) (self-reported) | 100 |
| Claw-Anything - CLI and GUI Tasks | 18 | Pass@1 (%) on the 50 tasks that need both the CLI and the GU | 100 |
| Embodied-BenchClaw Aerial-Bench - Interval Selection | 44 | Accuracy (%) on interval selection questions (counts, quanti | 100 |
| FirmCORe - Opportunity Detection | 74.51 | Macro-F1 (%) | 100 |
Interactive version: theaggregate.ai/model?slug=qwen-3-6-27b · How It Works · Data refreshed daily, snapshot 2026-10-20.