Compare AI models

Specifications
Parameters
141B—
Context window
64k128k
API pricing
Cheapest input
$2.00Mistral—
Cheapest output
$6.00Mistral—
$700Mistral · zdr—
Benchmarks
HumanEval
45.1%75.7%
MMLU
77.75%81.9%
GSM8K
78.6%90.9%
Overview
CompanyMistralZ.ai
Release dateApr 10 2024Jun 5 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

GLM-4-Air leads Mixtral 8×22B on 3 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). Mixtral 8×22B shipped 56 days before GLM-4-Air, so benchmark comparisons should account for the intervening progress.

Context windows are 64k (Mixtral 8×22B) vs 128k (GLM-4-Air). Mixtral 8×22B is open weight, while GLM-4-Air is closed.

On HumanEval, GLM-4-Air leads at 75.7% vs Mixtral 8×22B at 45.1%. On MMLU, GLM-4-Air leads at 81.9% vs Mixtral 8×22B at 77.75%. On GSM8K, GLM-4-Air leads at 90.9% vs Mixtral 8×22B at 78.6%.