NVIDIA models

NVIDIA publishes the Nemotron family of open models alongside its GPUs and NIM inference software. The Nemotron 3 family spans Nano, Super (120B, March 2026) and Ultra (550B, June 2026), with Super and Ultra using a hybrid Mamba-2/Transformer mixture-of-experts design, and Nemotron 3.5 Lightning followed in August 2026. NVIDIA hosts the models for testing on build.nvidia.com.

2 models trackedData checked October 10, 2026

Lineup

Model	Status	Weights	Tier	Context	Input / 1M	Output / 1M	Released
NVIDIA Nemotron 3 UltraNemotron 3	Current	Open weights	Advanced	1M	$0.50	$2.20	Jun 4, 2026
NVIDIA Nemotron 3.5 Lightning 30B-A3BNemotron 3.5	Current	Open weights	Compact	1M	$0.05	$0.20	Aug 11, 2026

The lowest-priced current model is NVIDIA Nemotron 3.5 Lightning 30B-A3B at $0.05 input and $0.20 output per million tokens. The most capable tier in the lineup is advanced, led by NVIDIA Nemotron 3 Ultra. The longest context window is 1M tokens (NVIDIA Nemotron 3 Ultra).

Frequently asked questions

What models does NVIDIA offer?

As of October 10, 2026, we track 2 NVIDIA models: NVIDIA Nemotron 3 Ultra, NVIDIA Nemotron 3.5 Lightning 30B-A3B. Status, prices and context windows for each are in the table on this page.

How do I access NVIDIA models?

Through NVIDIA's developer platform (build.nvidia.com), and for many models through cloud marketplaces and inference providers listed on each model page.

Does NVIDIA release open-weight models?

Nemotron 3 Ultra and Nemotron 3.5 Lightning use the OpenMDW 1.1 license; Nemotron 3 Super uses the NVIDIA Nemotron Open Model License.

Other providers

OpenAI Anthropic Google xAI Amazon (AWS)Meta Perplexity Mistral AI DeepSeek Alibaba (Qwen)Microsoft (Phi)Moonshot AI Z.ai (Zhipu AI)MiniMax IBM Ai2 (Allen Institute for AI)

Still choosing a model?

Send us your use case and constraints. We'll come back with a shortlist and an evaluation plan you can run on your own data.

Free and vendor-neutral: no model provider pays for placement.
Your details go only to partners you agree to.