Nemotron 3 Nano 30B A3B

NVIDIA

Free

Get

MLX Score2.6

PriceFreeno paid plan

RunsOn your Mac

DeveloperNVIDIA

LicenceOpenNVIDIA Nemotron Open Model License

CategoryModels

Mac14.0+macOS or later

Nemotron 3 Nano 30B A3B is a model NVIDIA trained from scratch for reasoning and non-reasoning tasks. It mixes Mamba-2 and mixture-of-experts layers with a few attention layers: 30 billion parameters, 3.5 billion active per token. A flag in the chat template turns the reasoning trace on or off.

A 4-bit MLX build is about 17 GB; on mlx.app's estimate it needs a Mac with 32 GB of memory.

Open weights under the NVIDIA Nemotron Open Model License, a custom licence. Run locally, nothing leaves your Mac.

How to Get It

Run it on your MacSee how

WebsiteVisit

Official downloads only. MLX never re-hosts an app, and never runs a command for you.

MLX Score

2.6out of 5

Hugging Face downloads and likesahead of 40%

Hugging Face: 1,715,687 downloads (30 days), 840 likes · read 2026-11-02

Ranked against the open models listings, from real public data only. How the score works.

Information

Developer: NVIDIA
Category: Open models
Price: Free (Open weights, free to download.)
Your data: Runs on your Mac
Licence: NVIDIA Nemotron Open Model License
Mac: macOS 14.0 or later
iPhone and iPad: —
Needs: 32 GB memory
Links: Website · Hugging Face
Facts checked: 2026-11-01

Sources

huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 · NVIDIA Nemotron Open Model License, hybrid Mamba-2/MoE, 30B total / 3.5B active, reasoning toggle, languages · checked 2026-11-01huggingface.co/mlx-community/NVIDIA-Nemotron-3-Nano-30B-A3B-4bit · 4-bit MLX build (about 17 GB) · checked 2026-11-01ml-explore.github.io/mlx/build/html/install.html · macOS 14.0 minimum (MLX needs Apple silicon and macOS 14.0+) · checked 2026-11-01mlx.app/run/nemotron-3-nano-30b-a3b · memory: smallest Mac (32 GB) where the 4-bit MLX build fits with room for context, from lib/compare (75% of unified memory usable) · checked 2026-11-01Icon from Hugging Face

See All