{"id":112139,"date":"2026-07-20T16:17:09","date_gmt":"2026-07-20T16:17:09","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/112139\/"},"modified":"2026-07-20T16:17:09","modified_gmt":"2026-07-20T16:17:09","slug":"nvidia-ai-super-agents-run-locally-on-dgx-station-desktops-with-ai-agent-toolkit","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/112139\/","title":{"rendered":"Nvidia AI super agents run locally on DGX Station desktops with AI agent toolkit"},"content":{"rendered":"<p>\tBecome a member of GB MAX to gain exclusive access to the industry and to the most influential global B2B leadership community in the business of gaming, entertainment, and tech. <a href=\"https:\/\/go.gamesbeat.com\/gb-max\/\" rel=\"nofollow noopener\" target=\"_blank\">Join now<\/a> and also get a VIP ticket to GamesBeat Next (Nov 2-3, SF).<\/p>\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.nvidia.com\" rel=\"nofollow noopener\" target=\"_blank\">Nvidia<\/a> said AI Agents are getting easier to use and creators can now run personal AI agents locally on DGX Station desktops.<\/p>\n<p class=\"wp-block-paragraph\">While DGX Stations are a lot more capable than your typical desktop computer, they are built in a desktop form factor, and now they can run AI agents locally, like a personal computer. <\/p>\n<p class=\"wp-block-paragraph\">In a post by Adel El Hallak, vice president of agentic AI at Nvidia, the AI and graphics chip maker announced at Siggraph that super agents have arrived on the desktop. <\/p>\n<p class=\"wp-block-paragraph\">El Hallak said <a href=\"https:\/\/blogs.nvidia.com\/blog\/siggraph-news-2026\/#nemoclaw-dgx-station\" rel=\"nofollow noopener\" target=\"_blank\">Nvidia DGX Station<\/a> is the ultimate deskside supercomputer for the AI era, and with Nvidia Agent Toolkit software, setup takes just three steps, and can be running in roughly 30 minutes.<\/p>\n<p class=\"wp-block-paragraph\">On DGX Station, Nvidia Agent Toolkit brings together Nvidia NemoClaw, the Nvidia Nemotron 3<br \/>Ultra open model, Nvidia Omniverse libraries as agent-accessible tools and skills, and a secure<br \/>runtime in a single local system \u2014 no internet required. <\/p>\n<p class=\"wp-block-paragraph\">As workloads scale, developers can connect multiple systems together to serve concurrent users, more agents and bigger models.<\/p>\n<p class=\"wp-block-paragraph\">This gives creatives and engineers the ability to own their own intelligence, with a system that<br \/>comes ready to run locally. The full stack\u2014 model, agent, tools \u2014 provides a platform for creating and running domain-specific \u201csuper agents\u201d that are customized with users\u2019 own data and knowledge.<\/p>\n<p class=\"wp-block-paragraph\">The open Nvidia Agent Toolkit stack on DGX Station includes:<\/p>\n<p class=\"wp-block-paragraph\">\u25cf Nvidia NemoClaw, open blueprints for building custom autonomous agents, packaging<br \/>the model, harness and runtime together as a starting point for teams building specialized,<br \/>domain-specific agents.<br \/>\u25cf Nvidia Nemotron 3 Ultra, a frontier 550-billion-parameter open model, is optimized to run<br \/>on DGX Station GB300 systems and serves as the model layer that teams can customize<br \/>for their own domains.<br \/>\u25cf Nvidia Omniverse libraries extend agent skills into physics simulation and 3D asset<br \/>workflows, giving creative and engineering professionals tools that go well beyond<br \/>general-purpose agent capabilities.<br \/>\u25cf NvidiaOpenShell, the open source secure runtime, keeps agents sandboxed and<br \/>governed according to defined policies for how agents interact with tools, systems and<br \/>data.<br \/>\u25cf Nvidia GB300 Grace Blackwell Ultra Desktop Superchip delivers data-center-level<br \/>performance from the desk on DGX Station, with up to 20 petaflops of FP4 AI compute<br \/>and 748GB of coherent memory to run large models such as Nemotron Ultra.<br \/>\u25cf Nvidia ConnectX-8 SuperNIC delivers up to 800GB\/s of bandwidth in DGX Station,<br \/>delivering extremely fast, efficient network connectivity, and supports linking up to two<br \/>DGX Stations to further scale model capacity and performance.<\/p>\n<p>Harness Efficiency at Scale<\/p>\n<p class=\"wp-block-paragraph\">For teams running agents at scale the economics shift fundamentally on DGX Station. Nemotron 3 Ultra, tuned for an open harness, delivers leading-edge performance without the per-token cost after the hardware purchase, so users build once and can run as much as they need.<\/p>\n<p class=\"wp-block-paragraph\">Nvidia has announced a blueprint for integrating Nvidia Omniverse libraries in Blender \u2014 giving NemoClaw agents callable RTX sensor simulation and physics tools to prepare 3D scenes for physical AI workflows.<\/p>\n<p class=\"wp-block-paragraph\">On DGX Station, designers and engineers can run the core pieces of that workflow \u2014 frontier<br \/>model, open harness, secure runtime, 3D tools \u2014 in one box, all connected and deployable<br \/>through an open blueprint. And Nvidia AI Agent tookit now has Omniverse libraries.<\/p>\n<p class=\"wp-block-paragraph\">Frontier models can orchestrate NemoClaw as a specialized sub-agent, delegating domain-<br \/>specific work to an agent running locally on DGX Station, with direct access to Omniverse tools and Blender.<\/p>\n<p class=\"wp-block-paragraph\">LangChain tuned its Deep Agents harness for Nemotron 3 Ultra, giving designers and engineers a production-ready path to benchmark-leading agentic performance at a fraction of the cost.<\/p>\n<p class=\"wp-block-paragraph\">Nous Research fine-tuned Nemotron 3 Ultra for its Hermes Agent harness and adopted it for production workloads \u2014 a direct demonstration of the value of owning intelligence. Tuning the<br \/>model for a developer\u2019s stack enables agents that are both faster and more capable for specific<br \/>domains. Hermes Agent has also added Blender to its Model Context Protocol catalog, letting<br \/>teams activate Blender directly from their agent \u2014 a live example of a tool-using NemoClaw  agent that can run on DGX Station.<\/p>\n<p class=\"wp-block-paragraph\">For teams running OpenClaw, this stack extends what\u2019s possible \u2014 bringing Nemotron 3 Ultra,<br \/>Omniverse tools and local inference on DGX Station into an environment where OpenClaw\u2019s<br \/>persistent, long-running agents can act on them continuously.<\/p>\n<p>Develop and Deploy Quickly With New Playbooks<\/p>\n<p class=\"wp-block-paragraph\">Two new playbooks are available now to help developers build and run agents out of the box with NemoClaw and dual-node deployments:<\/p>\n<p class=\"wp-block-paragraph\">\u25cf \u201cConnect Two DGX Stations for Distributed Workloads\u201d guides users on how to run powerful models with scalable performance connecting up to two DGX Station systems.<br \/>\u25cf \u201cRun NemoClaw With a Local LLM \/ On Dual DGX Station\u201d provides a walkthrough on how to set up NemoClaw on DGX Station, enabling out-of-the-box agents with frontier intelligence with Nemotron Ultra.<\/p>\n<p class=\"wp-block-paragraph\">Nvidia DGX Station is built and available to order from ASUS, Dell Technologies, Exxact,<br \/>GIGABYTE, HP, MSI and Supermicro.<\/p>\n","protected":false},"excerpt":{"rendered":"Become a member of GB MAX to gain exclusive access to the industry and to the most influential&hellip;\n","protected":false},"author":2,"featured_media":112140,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[57567,179,7493,405,57568,58,57569,50656],"class_list":["post-112139","post","type-post","status-publish","format-standard","has-post-thumbnail","category-agentic-ai","tag-adel-el-hallak","tag-agentic-ai","tag-agentic-artificial-intelligence","tag-ai-agents","tag-dgx-stations","tag-nvidia","tag-nvidia-agent-toolkit","tag-siggraph"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/112139","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=112139"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/112139\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/112140"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=112139"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=112139"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=112139"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}