Compare AI models

Specifications
Context window
32k128k
Benchmarks
HumanEval
45.1%75.7%
MMLU
81.2%81.9%
GSM8K
81%90.9%
Overview
CompanyMistralZ.ai
Release dateFeb 26 2024Jun 5 2024
AccessClosedClosed
Model detailsView modelView model

Frequently asked questions

GLM-4-Air leads Mistral Large on 3 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). Mistral Large shipped 100 days before GLM-4-Air, so benchmark comparisons should account for the intervening progress.

Context windows are 32k (Mistral Large) vs 128k (GLM-4-Air).

On HumanEval, GLM-4-Air leads at 75.7% vs Mistral Large at 45.1%. On MMLU, GLM-4-Air leads at 81.9% vs Mistral Large at 81.2%. On GSM8K, GLM-4-Air leads at 90.9% vs Mistral Large at 81%.