Compare AI models

Specifications
Parameters
284B16B
Context window
1M8k
API pricing
Cheapest input
$0.05Baidu—
Cheapest output
$0.10Baidu—
$15.00Baidu · fp8—
Benchmarks
LiveCodeBench
91.6%14.3%
GPQA Diamond
88.1%35.2%
Overview
CompanyDeepSeekMoonshot AI
Release dateApr 24 2026Feb 22 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-V4-Flash leads Moonlight on 2 of the 2 benchmarks they both report (LiveCodeBench, GPQA Diamond). Moonlight shipped 426 days before DeepSeek-V4-Flash, so benchmark comparisons should account for the intervening progress.

DeepSeek-V4-Flash has 284B parameters, while Moonlight has 16B. Context windows are 1M (DeepSeek-V4-Flash) vs 8k (Moonlight).

On LiveCodeBench, DeepSeek-V4-Flash leads at 91.6% vs Moonlight at 14.3%. On GPQA Diamond, DeepSeek-V4-Flash leads at 88.1% vs Moonlight at 35.2%.