Compare AI models

Specifications
Parameters
16B—
Context window
8k200k
Benchmarks
HumanEval
48.1%97.3%
GPQA Diamond
35.2%77.6%
SimpleQA
5.6%20.2%
MMLU
70%90%
Overview
CompanyMoonshot AIOpenAI
Release dateFeb 22 2025Apr 16 2025
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

o4-mini leads Moonlight on 4 of the 4 benchmarks they both report (HumanEval, GPQA Diamond, SimpleQA, MMLU). Moonlight shipped 53 days before o4-mini, so benchmark comparisons should account for the intervening progress.

Context windows are 8k (Moonlight) vs 200k (o4-mini). Moonlight is open weight, while o4-mini is closed.

On HumanEval, o4-mini leads at 97.3% vs Moonlight at 48.1%. On GPQA Diamond, o4-mini leads at 77.6% vs Moonlight at 35.2%. On SimpleQA, o4-mini leads at 20.2% vs Moonlight at 5.6%. On MMLU, o4-mini leads at 90% vs Moonlight at 70%.