Compare AI models

Specifications
Parameters
7B16B
Context window
32k8k
Benchmarks
HumanEval
38.4%48.1%
MATH
13.2%45.3%
Overview
CompanyMistralMoonshot AI
Release dateMay 22 2024Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Moonlight leads Mistral 7B v0.3 on 2 of the 2 benchmarks they both report (HumanEval, MATH). Mistral 7B v0.3 shipped 276 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Mistral 7B v0.3 has 7B parameters, while Moonlight has 16B. Context windows are 32k (Mistral 7B v0.3) vs 8k (Moonlight).

On HumanEval, Moonlight leads at 48.1% vs Mistral 7B v0.3 at 38.4%. On MATH, Moonlight leads at 45.3% vs Mistral 7B v0.3 at 13.2%.