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

NVIDIA has launched Nemotron 3 Nano Omni, an open multimodal model that unifies vision, audio, and language capabilities in a single 30B-A3B hybrid mixture-of-experts architecture. Unlike agentic systems that chain separate models for each modality, this model eliminates the need for separate perception models, delivering up to 9x higher throughput than comparable open omni models. It tops six leaderboards for document intelligence, video, and audio understanding. Use cases include computer use agents, document intelligence, and audio-video reasoning. The model is available with open weights on Hugging Face, OpenRouter, and NVIDIA's NIM microservice platform, and is already being adopted by companies including H Company, Palantir, and Foxconn.