← All models
N

NVIDIA: Nemotron 3 Super

nvidia/nemotron-3-super-120b-a12b

Tool useJSONReasoningStreaming

NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. Built on a hybrid Mamba-Transformer...

Pricing

Input

$0.09 / 1M

Output

$0.45 / 1M

Specs

Context

1,000,000 tokens

Input

text

Output

text

Released: 2026-03

Supported parameters

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstoptemperaturetool_choicetoolstop_ktop_logprobstop_p

Open weights · HuggingFace

436,835 downloads/mo
251 likes
other text-generation
View on HuggingFace →