Developers wanting one small open model that reads video, audio, images and text together should test it before stitching models together.

Kenning's take

What it does

Nemotron 3 Nano Omni is an open multimodal model from NVIDIA that handles video, audio, images and text in one system, so agents do not have to pass data between separate models. It is a 30B mixture-of-experts model with about 3B active parameters and a 256K context window, aimed at tasks such as computer use, document reading and audio-video reasoning. It is available on Hugging Face, OpenRouter and build.nvidia.com.

• Video, audio, image and text input in one model
• 30B-A3B hybrid mixture-of-experts
• 256K token context window
• Computer use and document intelligence
• Available as an NVIDIA NIM microservice

Who it suits. Developers building agents that need to understand screens, documents, speech or video.

Key facts

Category

AI models

Part of

NVIDIA Nemotron 3

Made by

NVIDIA

Based in

United States

Released

28 April 2026

How to get it

open weights on Hugging Face; hosted on OpenRouter, build.nvidia.com and 25+ partner platforms

Works on

API, self-hosted

In the UK

Yes

Free plan

free open weights

API price

set by third-party providers

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.