Compare AI models

Specifications
Parameters
3B—
Context window
128k128k
Benchmarks
HumanEval
77.4%75.7%
MATH
51.7%57.9%
Overview
CompanyMistralZ.ai
Release dateOct 16 2024Jun 5 2024
AccessClosedClosed
Model detailsView modelView model

Frequently asked questions

Ministral 3B and GLM-4-Air are evenly matched across the 2 benchmarks they both report (HumanEval, MATH). GLM-4-Air shipped 133 days before Ministral 3B, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Ministral 3B) vs 128k (GLM-4-Air).

On HumanEval, Ministral 3B leads at 77.4% vs GLM-4-Air at 75.7%. On MATH, GLM-4-Air leads at 57.9% vs Ministral 3B at 51.7%.