Qwen-7B — benchmark results

Provider: Alibaba. Released 2023-08-01. Access: Open.

Unified ELO 1302 ± 14, rank #1531 of 1776 rated models, from 28 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
ARC Challenge (AI2)75.3Accuracy (%)70.5
T-Eval59.5Overall Score (%)70
GSM8K51.7Accuracy (%)52.1
CyberMetric52.9Accuracy (%)41.7
CMMLU58.665-shot Avg Accuracy (%)40
Big-Bench Hard45Average (%)39.6
BoolQ76.4Accuracy (%)38.3
VMLU32.81Average (%)37.5
VMLU - Humanities34.15Accuracy (%)37.5
VMLU - Other32.68Accuracy (%)37.5
VMLU - STEM30.64Accuracy (%)37.5
InfiBench31.69Score (%)35.2

Interactive version: theaggregate.ai/model?slug=qwen-7b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.