Nemotron 3 Nano Omni

Nemotron 3 Nano Omni is a multimodal large language model from NVIDIA built with a hybrid Mixture-of-Experts (MoE) architecture.

Use this snap to quickly install an optimized environment for local inference with Nemotron 3 Nano Omni.

The snap includes the following hardware-optimized inference engines:

• cpu: Optimized for x64 and ARM (armv8, armv9) CPUs
• nvidia-gpu: CUDA-enabled GPU acceleration

The most suitable engine is automatically selected based on the available hardware.

Install:

sudo snap install nemotron-3-nano-omni

Run:

nemotron-3-nano-omni

Some accelerators require extra drivers to be usable with this snap: https://documentation.ubuntu.com/inference-snaps/how-to/setup/drivers/

License:

The Nemotron 3 Nano Omni model is provided by NVIDIA Corporation under the NVIDIA Open Model License: https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/

The licenses of all the bundled software can be found after installing the snap at /snap/nemotron-3-nano-omni/current/usr/share/doc.