Compare AI models

Specifications
Parameters
671B30B
Context window
128k128k
API pricing
Cheapest input
—$0.30DeepInfra
Cheapest output
—$1.20DeepInfra
—$120ModelRun · bf16
Benchmarks
Humanity's Last Exam · text only, no tools
30.6%22%
GPQA Diamond
85.7%83.5%
Overview
CompanyDeepSeekMeta
Release dateDec 1 2025Aug 10 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-V3.2-Speciale leads Muse Glimmer on 2 of the 2 benchmarks they both report (Humanity's Last Exam, GPQA Diamond). DeepSeek-V3.2-Speciale shipped 252 days before Muse Glimmer, so benchmark comparisons should account for the intervening progress.

DeepSeek-V3.2-Speciale has 671B parameters, while Muse Glimmer has 30B. Context windows are 128k (DeepSeek-V3.2-Speciale) vs 128k (Muse Glimmer).

On Humanity's Last Exam · text only, no tools, DeepSeek-V3.2-Speciale leads at 30.6% vs Muse Glimmer at 22%. On GPQA Diamond, DeepSeek-V3.2-Speciale leads at 85.7% vs Muse Glimmer at 83.5%.