Compare AI models

Specifications
Parameters
7B16B
Context window
8k8k
Benchmarks
HumanEval
26.2%48.1%
WinoGrande
74.2%74.6%
MMLU
62.5%70%
GSM8K
36.5%77.4%
Overview
CompanyMistralMoonshot AI
Release dateSep 27 2023Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Moonlight leads Mistral 7B on 4 of the 4 benchmarks they both report (HumanEval, WinoGrande, MMLU, GSM8K). Mistral 7B shipped 514 days before Moonlight, so benchmark comparisons should account for the intervening progress.

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

On HumanEval, Moonlight leads at 48.1% vs Mistral 7B at 26.2%. On WinoGrande, Moonlight leads at 74.6% vs Mistral 7B at 74.2%. On MMLU, Moonlight leads at 70% vs Mistral 7B at 62.5%. On GSM8K, Moonlight leads at 77.4% vs Mistral 7B at 36.5%.