Compare AI models

Specifications
Parameters
—16B
Context window
2M8k
Benchmarks
LiveCodeBench
36%14.3%
MATH
91.8%45.3%
GPQA Diamond
64.7%35.2%
SimpleQA
44.3%5.6%
MMLU-Pro
79.1%42.4%
Overview
CompanyGoogleMoonshot AI
Release dateFeb 5 2025Feb 22 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemini 2.0 Pro leads Moonlight on 5 of the 5 benchmarks they both report. Gemini 2.0 Pro shipped 17 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Context windows are 2M (Gemini 2.0 Pro) vs 8k (Moonlight). Gemini 2.0 Pro is closed, while Moonlight is open weight.

On LiveCodeBench, Gemini 2.0 Pro leads at 36% vs Moonlight at 14.3%. On MATH, Gemini 2.0 Pro leads at 91.8% vs Moonlight at 45.3%. On GPQA Diamond, Gemini 2.0 Pro leads at 64.7% vs Moonlight at 35.2%. On SimpleQA, Gemini 2.0 Pro leads at 44.3% vs Moonlight at 5.6%. On MMLU-Pro, Gemini 2.0 Pro leads at 79.1% vs Moonlight at 42.4%.