NVIDIA Launches Nemotron 3 Nano Omni Model, Unifying Vision, Audio and Language for up to 9x More Efficient AI Agents

NVIDIA released Nemotron 3 Nano Omni, an open-source multimodal AI model combining vision, audio, image and text processing in a single system. The 30B-A3B hybrid mixture-of-experts model achieves 9x higher throughput than comparable open multimodal models while maintaining leading accuracy across document intelligence, video and audio tasks. The model is now available for enterprises and developers building efficient agentic systems with deployment flexibility across local, data center and cloud environments.