Compare AI models

Specifications
Parameters
16B—
Context window
8k1M
Benchmarks
LiveCodeBench
14.3%37.8%
HumanEval
48.1%86.6%
MATH
45.3%81.1%
MMLU-Pro
42.4%64.5%
GSM8K
77.4%93.8%
Overview
CompanyMoonshot AIQwen
Release dateFeb 22 2025Nov 15 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Qwen2.5-Turbo (1M) leads Moonlight on 5 of the 5 benchmarks they both report. Qwen2.5-Turbo (1M) shipped 99 days before Moonlight, so benchmark comparisons should account for the intervening progress.

Context windows are 8k (Moonlight) vs 1M (Qwen2.5-Turbo (1M)). Moonlight is open weight, while Qwen2.5-Turbo (1M) is closed.

On LiveCodeBench, Qwen2.5-Turbo (1M) leads at 37.8% vs Moonlight at 14.3%. On HumanEval, Qwen2.5-Turbo (1M) leads at 86.6% vs Moonlight at 48.1%. On MATH, Qwen2.5-Turbo (1M) leads at 81.1% vs Moonlight at 45.3%. On MMLU-Pro, Qwen2.5-Turbo (1M) leads at 64.5% vs Moonlight at 42.4%. On GSM8K, Qwen2.5-Turbo (1M) leads at 93.8% vs Moonlight at 77.4%.