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.
Unified ELO 1650 ± 1, rank #128 of 1392 rated models, from 250 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CheXpercept | 92.2 | Stage 1 (End-to-End) (self-reported) | 100 |
| LLM Stats (EmbSpatialBench) | 84.6 | Score (%) | 100 |
| Nejumi 4 - GLP - Information Retrieval | 82.46 | Score (%) | 100 |
| RefSpatialBench | 70 | RefSpatialBench (self-reported) | 100 |
| The Age of Curiosity Meets the Age of AI: Benc | 4.98 | Total (self-reported) | 100 |
| MERA - ruHHH | 93.82 | Accuracy (%) | 99.8 |
| MERA - ruModAr | 99.98 | EM (%) | 99.5 |
| MERA - ruHateSpeech | 93.21 | Accuracy (%) | 98.6 |
| MERA - MathLogicQA | 99.65 | Accuracy (%) | 97.4 |
| MERA - ruOpenBookQA | 96.5 | Accuracy (%) | 97.4 |
| MERA - ruCodeEval | 77.2 | pass@1 (%) | 97.1 |
| MERA - ruHumanEval | 79.51 | pass@1 (%) | 97.1 |
Interactive version: theaggregate.ai/model?slug=qwen-3-6-27b · How It Works · Data refreshed daily, snapshot 2026-09-05.