Compare AI models

Specifications
Parameters
0.5B/1.5B/7B/57B/72B—
Context window
128k128k
Benchmarks
HumanEval
86%75.7%
MATH
69%57.9%
MMLU
82.3%81.9%
GSM8K
91.1%90.9%
Overview
CompanyQwenZ.ai
Release dateJun 6 2024Jun 5 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Qwen2 leads GLM-4-Air on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). GLM-4-Air shipped 1 day before Qwen2, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Qwen2) vs 128k (GLM-4-Air). Qwen2 is open weight, while GLM-4-Air is closed.

On HumanEval, Qwen2 leads at 86% vs GLM-4-Air at 75.7%. On MATH, Qwen2 leads at 69% vs GLM-4-Air at 57.9%. On MMLU, Qwen2 leads at 82.3% vs GLM-4-Air at 81.9%. On GSM8K, Qwen2 leads at 91.1% vs GLM-4-Air at 90.9%.