Compare AI models

Specifications
Parameters
2B/7B—
Context window
8k128k
Benchmarks
HumanEval
32.3%75.7%
MATH
24.3%57.9%
MMLU
64.3%81.9%
GSM8K
46.4%90.9%
Overview
CompanyGoogleZ.ai
Release dateFeb 21 2024Jun 5 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

GLM-4-Air leads Gemma 1 on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). Gemma 1 shipped 105 days before GLM-4-Air, so benchmark comparisons should account for the intervening progress.

Context windows are 8k (Gemma 1) vs 128k (GLM-4-Air). Gemma 1 is open weight, while GLM-4-Air is closed.

On HumanEval, GLM-4-Air leads at 75.7% vs Gemma 1 at 32.3%. On MATH, GLM-4-Air leads at 57.9% vs Gemma 1 at 24.3%. On MMLU, GLM-4-Air leads at 81.9% vs Gemma 1 at 64.3%. On GSM8K, GLM-4-Air leads at 90.9% vs Gemma 1 at 46.4%.