Compare AI models

Specifications
Parameters
2B/7B—
Context window
8k—
Benchmarks
HumanEval
32.3%72%
MATH
24.3%47.9%
MMLU
64.3%81.5%
GSM8K
46.4%87.6%
Overview
CompanyGoogleZ.ai
Release dateFeb 21 2024Jan 16 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

GLM-4 leads Gemma 1 on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). GLM-4 shipped 36 days before Gemma 1, so benchmark comparisons should account for the intervening progress.

Gemma 1 is open weight, while GLM-4 is closed.

On HumanEval, GLM-4 leads at 72% vs Gemma 1 at 32.3%. On MATH, GLM-4 leads at 47.9% vs Gemma 1 at 24.3%. On MMLU, GLM-4 leads at 81.5% vs Gemma 1 at 64.3%. On GSM8K, GLM-4 leads at 87.6% vs Gemma 1 at 46.4%.