Compare AI models

Specifications
Parameters
671B—
Context window
128k—
API pricing
Cheapest input
$0.27SiliconFlow—
Cheapest output
$0.41SiliconFlow—
$74.50SiliconFlow · fp8—
Benchmarks
SWE-Bench Verified
67.8%77.4%
GPQA Diamond
79.9%89.5%
Overview
CompanyDeepSeekMeta
Release dateSep 29 2025Apr 8 2026
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Muse Spark leads DeepSeek V3.2 Exp on 2 of the 2 benchmarks they both report (SWE-Bench Verified, GPQA Diamond). DeepSeek V3.2 Exp shipped 191 days before Muse Spark, so benchmark comparisons should account for the intervening progress.

DeepSeek V3.2 Exp is open weight, while Muse Spark is closed.

On SWE-Bench Verified, Muse Spark leads at 77.4% vs DeepSeek V3.2 Exp at 67.8%. On GPQA Diamond, Muse Spark leads at 89.5% vs DeepSeek V3.2 Exp at 79.9%.