Compare AI models

Specifications
Parameters
671B16B
Context window
128k8k
API pricing
Cheapest input
$0.50DeepInfra—
Cheapest output
$2.15DeepInfra—
$208DeepInfra · fp4—
Benchmarks
LiveCodeBench
73.3%14.3%
GPQA Diamond
81%35.2%
SimpleQA
27.8%5.6%
MMLU-Pro
85%42.4%
Overview
CompanyDeepSeekMoonshot AI
Release dateMay 28 2025Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek R1-0528 leads Moonlight on 4 of the 4 benchmarks they both report (LiveCodeBench, GPQA Diamond, SimpleQA, MMLU-Pro). Moonlight shipped 95 days before DeepSeek R1-0528, so benchmark comparisons should account for the intervening progress.

DeepSeek R1-0528 has 671B parameters, while Moonlight has 16B. Context windows are 128k (DeepSeek R1-0528) vs 8k (Moonlight).

On LiveCodeBench, DeepSeek R1-0528 leads at 73.3% vs Moonlight at 14.3%. On GPQA Diamond, DeepSeek R1-0528 leads at 81% vs Moonlight at 35.2%. On SimpleQA, DeepSeek R1-0528 leads at 27.8% vs Moonlight at 5.6%. On MMLU-Pro, DeepSeek R1-0528 leads at 85% vs Moonlight at 42.4%.