Nvidia Corp. today launched a powerful\u00a0reasoning artificial intelligence model that unifies text,\u00a0vision\u00a0and speech, capable of acting as the \u201cbrains\u201d of faster, smarter agentic AI applications.\u00a0<\/p>\n
Dubbed\u00a0Nemotron\u00a03 Nano Omni<\/a>, and weighing in at about 30 billion parameters, the new\u00a0state-of-the-art\u00a0model uses mixture-of-experts\u00a0architecture to deliver extremely low latency and\u00a0provides high flexibility and control.\u00a0<\/p>\n Nvidia combined vision and audio encoders with its 30B-AD3B hybrid\u00a0MoE\u00a0architecture to\u00a0eliminate\u00a0the need for separate perception modules, allowing its AI model to unify\u00a0everything into one. The company said this allowed the model to improve efficiency at scale and provide up to nine times faster throughput than other open omni models on the market.\u00a0<\/p>\n \u201cTo build useful agents, you can\u2019t wait seconds for a model to interpret a screen,\u201d said Gautier\u00a0Cloix,\u00a0chief\u00a0executive of H Company. \u201cBy building on\u00a0Nemotron\u00a03 Nano Omni, our agents can rapidly interpret full HD screen recordings \u2014 something that\u00a0wasn\u2019t\u00a0practical before.\u201d\u00a0<\/p>\n The result is a lower cost and higher scalability. With its smaller size,\u00a0it can also be compressed enough to run on higher-end consumer hardware and execute efficiently on enterprise cloud deployments.\u00a0<\/p>\n The company said it is designed to run alongside other proprietary cloud models or other Nvidia\u00a0Nemotron\u00a0open models, such as\u00a0Nemotron\u00a03 Super\u00a0for high-frequency execution\u00a0or Super for complex planning.\u00a0<\/p>\n The new model allows for rapid understanding of documents, computer displays, voice activity,\u00a0video\u00a0and more. This makes it the suitable interface for working with people and bridging to more complex machine states. It can take conversational replies from a user and quickly turn it around into reasoning.\u00a0<\/p>\n Nvidia said the\u00a0Nemotron\u00a0family \u2013 including Ultra, Super and Nano \u2013 has seen over 50 million downloads in the past year. The Omni variant extends the family\u2019s capabilities into the multimodal and agentic domains.\u00a0<\/p>\n The new model is now available on Hugging Face,\u00a0OpenRouter\u00a0and\u00a0build.nvidia.com<\/a>\u00a0as an\u00a0Nvidia NIM microservice<\/a>. As an open, lightweight model, it\u2019s also designed for developers to build on\u00a0and\u00a0deploy on\u00a0local hardware, including\u00a0the\u00a0Nvidia DGX Spark<\/a>\u00a0and other hardware.\u00a0<\/p>\n Image: Nvidia<\/p>\n Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE\u2019s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.<\/p>\n 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more About SiliconANGLE Media<\/p>\n SiliconANGLE Media is a recognized leader in digital media innovation, uniting breakthrough technology, strategic insights and real-time audience engagement. As the parent company of SiliconANGLE<\/a>, theCUBE Network<\/a>, theCUBE Research<\/a>, CUBE365<\/a>, theCUBE AI<\/a> and theCUBE SuperStudios \u2014 with flagship locations in Silicon Valley and the New York Stock Exchange \u2014 SiliconANGLE Media operates at the intersection of media, technology and AI.<\/p>\n Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.<\/p>\n","protected":false},"excerpt":{"rendered":"Nvidia Corp. today launched a powerful\u00a0reasoning artificial intelligence model that unifies text,\u00a0vision\u00a0and speech, capable of acting as the…\n","protected":false},"author":2,"featured_media":20926,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[179,7493,999,14658,396],"class_list":{"0":"post-20925","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-agentic-ai","8":"tag-agentic-ai","9":"tag-agentic-artificial-intelligence","10":"tag-kyt-dotson","11":"tag-nvidia-introduces-nemotron-3-nano-omni-with-vision-and-speech-for-powerful-agentic-ai-use","12":"tag-siliconangle"},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/20925","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=20925"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/20925\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/20926"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=20925"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=20925"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=20925"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
\n11.4k+ theCUBE alumni \u2014 Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network.<\/p>\n