Compare AI models

Specifications
Parameters
47B16B
Context window
32k8k
Benchmarks
HumanEval
40.2%48.1%
WinoGrande
77.2%74.6%
MMLU
70.6%70%
GSM8K
58.4%77.4%
Overview
CompanyMistralMoonshot AI
Release dateDec 8 2023Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Mixtral 8×7B and Moonlight are evenly matched across the 4 benchmarks they both report (HumanEval, WinoGrande, MMLU, GSM8K). Mixtral 8×7B shipped 442 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Mixtral 8×7B has 47B parameters, while Moonlight has 16B. Context windows are 32k (Mixtral 8×7B) vs 8k (Moonlight).

On HumanEval, Moonlight leads at 48.1% vs Mixtral 8×7B at 40.2%. On WinoGrande, Mixtral 8×7B leads at 77.2% vs Moonlight at 74.6%. On MMLU, Mixtral 8×7B leads at 70.6% vs Moonlight at 70%. On GSM8K, Moonlight leads at 77.4% vs Mixtral 8×7B at 58.4%.