Compare AI models

Specifications
Parameters
22B16B
Context window
32k8k
Benchmarks
LiveCodeBench
31.5%14.3%
HumanEval
81.1%48.1%
Overview
CompanyMistralMoonshot AI
Release dateMay 29 2024Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Codestral 22B leads Moonlight on 2 of the 2 benchmarks they both report (LiveCodeBench, HumanEval). Codestral 22B shipped 269 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Codestral 22B has 22B parameters, while Moonlight has 16B. Context windows are 32k (Codestral 22B) vs 8k (Moonlight).

On LiveCodeBench, Codestral 22B leads at 31.5% vs Moonlight at 14.3%. On HumanEval, Codestral 22B leads at 81.1% vs Moonlight at 48.1%.