Compare AI models

Specifications
Parameters
141B1.5B/7B/72B
Context window
64k4k
API pricing
Cheapest input
$2.00Mistral—
Cheapest output
$6.00Mistral—
$700Mistral · zdr—
Benchmarks
GSM8K
78.6%96.7%
Overview
CompanyMistralQwen
Release dateApr 10 2024Aug 8 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen2-Math leads Mixtral 8×22B on 1 of the 1 benchmark they both report (GSM8K). Mixtral 8×22B shipped 120 days before Qwen2-Math, so benchmark comparisons should account for the intervening progress.

Mixtral 8×22B has 141B parameters, while Qwen2-Math has 1.5B/7B/72B. Context windows are 64k (Mixtral 8×22B) vs 4k (Qwen2-Math).

On GSM8K, Qwen2-Math leads at 96.7% vs Mixtral 8×22B at 78.6%.