NVIDIA has launched Nemotron 3 Nano Omni, an open multimodal reasoning model that combines vision, audio, and language capabilities into a single system, delivering up to nine times higher throughput than comparable open omni models. The model is designed to serve as the perception layer in agentic AI systems, enabling faster and more accurate responses across video, audio, image, text, documents, charts, and graphical interfaces.
Most agentic AI systems today rely on separate models for vision, speech, and language, creating latency through repeated inference passes, fragmented context across modalities, and compounding costs and inaccuracies over time. Nemotron 3 Nano Omni eliminates this fragmentation through a 30 billion parameter hybrid mixture-of-experts architecture with integrated vision and audio encoders, allowing a single model to handle the full perception workload. The model tops six leaderboards for complex document intelligence and video and audio understanding, and supports a 256,000 token context window.
Key use cases include computer use agents navigating graphical interfaces, document intelligence for enterprise analysis and compliance workflows, and audio-video understanding for customer service and research applications. The model is available with open weights, datasets, and training techniques, and can be deployed across local systems, data centers, and cloud environments to meet regulatory, sovereignty, or data localization requirements. It is available immediately via Hugging Face, OpenRouter, build.nvidia.com, and more than 25 partner platforms. Early adopters include Aible, Foxconn, Palantir, and H Company, with Dell Technologies, DocuSign, Infosys, and Oracle among those evaluating the model. The broader Nemotron 3 model family has seen more than 50 million downloads in the past year.
KEY QUOTE:
“To build useful agents, you can’t wait seconds for a model to interpret a screen. By building on Nemotron 3 Nano Omni, our agents can rapidly interpret full HD screen recordings — something that wasn’t practical before. This isn’t just a speed boost: it’s a fundamental shift in how our agents perceive and interact with digital environments in real time.”
Gautier Cloix, CEO, H Company
