Compare AI models

Specifications
Parameters
16B1.5B/7B/72B
Context window
8k4k
Benchmarks
MATH
45.3%85.9%
GSM8K
77.4%95.9%
Overview
CompanyMoonshot AIQwen
Release dateFeb 22 2025Sep 19 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen2.5-Math leads Moonlight on 2 of the 2 benchmarks they both report (MATH, GSM8K). Qwen2.5-Math shipped 156 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Moonlight has 16B parameters, while Qwen2.5-Math has 1.5B/7B/72B. Context windows are 8k (Moonlight) vs 4k (Qwen2.5-Math).

On MATH, Qwen2.5-Math leads at 85.9% vs Moonlight at 45.3%. On GSM8K, Qwen2.5-Math leads at 95.9% vs Moonlight at 77.4%.