Compare AI models

Specifications
Parameters
22B14.3B
Context window
32k32k
Benchmarks
HumanEval
81.1%34.2%
Overview
CompanyMistralQwen
Release dateMay 29 2024Mar 28 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Codestral 22B leads Qwen1.5-MoE on 1 of the 1 benchmark they both report (HumanEval). Qwen1.5-MoE shipped 62 days before Codestral 22B, so benchmark comparisons should account for the intervening progress.

Codestral 22B has 22B parameters, while Qwen1.5-MoE has 14.3B. Context windows are 32k (Codestral 22B) vs 32k (Qwen1.5-MoE).

On HumanEval, Codestral 22B leads at 81.1% vs Qwen1.5-MoE at 34.2%.