Qwen 3.8 27B (Low): benchmark results

Provider: Alibaba. Released 2026-08-14. Access: Open.

Unified ELO 1628 ± 1, rank #327 of 1761 rated models, from 19 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Artificial Analysis Intelligence Index33.76Intelligence Index86.1
AA Long Context Reasoning77.33Accuracy (%)81.7
AA GDPval1407.51ELO77.6
AA GPQA Diamond84.55Accuracy (%)76.6
Tau3 Banking32.16Success Rate (%)72
AA Humanity's Last Exam14.04Accuracy (%)64.7
AA MMMU-Pro73.76Accuracy (%)63.8
AA Omniscience-26.67Score59.3
AA Omniscience - Software Engineering (SWE)23.5Accuracy (%)44
AA Omniscience - Science, Engineering & Mathematics25.9Accuracy (%)40.5
AA Omniscience - Health17.7Accuracy (%)38.3
AA-Omniscience Accuracy17.13Accuracy (%)35.9

Interactive version: theaggregate.ai/model?slug=qwen-3-8-27b-low · How It Works · Data refreshed daily, snapshot 2026-09-05.