NVIDIA Nemotron 3 Nano Omni delivers long-context omni-modal intelligence
nvidia released nemotron 3 nano omni, an open omni-modal model for document, audio, and video agents. it combines a mamba-transformer moe backbone with new encoders to handle 100+ page documents and long-form video. benchmarks show it leads mm-longbench-doc, voicesense, and worldsense while delivering up to 9.2x higher system efficiency than alternatives.
NVIDIA released Nemotron 3 Nano Omni, a new open model extending the Nemotron line to support text, image, video, and audio inputs. The architecture pairs a Nemotron 3 hybrid Mamba-Transformer Mixture-of-Experts backbone with a C-RADIOv4-H vision encoder and a Parakeet-TDT-0.6B-v2 audio encoder. It targets five specific enterprise workloads: complex document analysis (capable of handling 100+ page files), automatic speech recognition, long audio-video understanding, agentic computer use, and general multimodal reasoning.
NVIDIA claims the model achieves top accuracy on leaderboards including MMlongbench-Doc, OCRBenchV2, WorldSense, and VoiceBench. Efficiency metrics indicate Nemotron 3 Nano Omni delivers up to 9x higher throughput and 2.9x faster single-stream reasoning speeds on multimodal use cases compared to alternatives, while ranking as the most cost-efficient open video understanding model on MediaPerf. BF16, FP8, and NVFP4 checkpoints are available on Hugging Face.