Qwen 3.8 27B: benchmark results

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

Unified ELO 1689 ± 1, rank #70 of 1392 rated models, from 137 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
LiveBench Python75Score98.1
AI for Education Pedagogy - Science93.44Accuracy (%)96
OpenRouter Tau2-Bench Airline78.7Accuracy (%)95.8
OSWorld-Verified84.3Success rate (self-reported)95.4
Bullshit Benchmark76.4BS Detection Rate (%)93.6
LLM Stats Score45.17LLM Stats Score (conservative rating)91.7
RealWorldQA85.9RealWorldQA (self-reported)91.7
BenchmarkList ECI143.19Capability Index (ECI)91.5
WebDev Arena1594.78Arena Score91.5
LLM Stats (MathVision)94.6Score (%)91.2
LLM Stats (CharXiv-R)90.2Score (%)90.7
WebDev Arena (Reference-Based Design)1614Arena Score89.3

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