Qwen 3.8 27B (Medium): benchmark results

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

Unified ELO 1641 ± 1, rank #279 of 1761 rated models, from 19 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Tau3 Banking47.42Success Rate (%)97
AA Long Context Reasoning79.67Accuracy (%)88.1
Artificial Analysis Intelligence Index35.18Intelligence Index87.1
AA GDPval1434.2ELO81
AA GPQA Diamond84.55Accuracy (%)76.6
AA MMMU-Pro74.16Accuracy (%)66.3
AA Humanity's Last Exam14.09Accuracy (%)65
AA Omniscience-36.15Score51.1
AA Omniscience - Science, Engineering & Mathematics27Accuracy (%)47.3
AA Omniscience - Health19Accuracy (%)44.4
AA Omniscience - Software Engineering (SWE)23.4Accuracy (%)43.6
AA-Omniscience Accuracy18.35Accuracy (%)42.3

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