NVIDIA Nemotron and Cosmos Reason Models Empower Enterprise AI Agents and Physical AI for Industry L

NVIDIA Expands NEMOtron and Cosmos Model Families with Advanced Reasoning Capabilities, Adopted by Industry Leaders Including CrowdStrike, Uber, and Zoom

AI Agents are projected to generate up to $450 billion through revenue gAIns and cost savings by 2028, according to CapGemini. To build these Agents, developers are increASIngly turning to higher-performing reasoning models to enhance AI Agent platforms and Physical AI systems.

At SIGGRAPH, Nvidia announced the expansion of two model families with reasoning capabilities—NVIDIA Nemotron and NVIDIA Cosmos—which are being adopted by industry leaders to boost Productivity through teams of AI agents and humanoid robots. Enterprises leverAGIng these model families include CrowdStrike, Uber, Magna, NetAPP, and Zoom.

Nemotron: Highest Accuracy and Efficiency for Agentic Enterprise AI

As enterprises develop AI agents to handle complex, multistep tasks, models that deliver strong reasoning accuracy with efficient token generation are essential for Intelligent, autonomous decision-making at scale. NVIDIA Nemotron is a family of advanced open reasoning models that combine leading models, NVIDIA-curated open datasets, and cutting-edge AI techniques to provide an accurate and efficient foundation for AI Agents.

The latest Nemotron models achieve leading efficiency through three innovations: a new hybrid model architecture, compACT quantized models, and a configurable thinking budget that gives developers control over token generation, resulting in 60% lower reasoning costs. This combination allows the models to reason more deeply and respond faster without requiring additional time or computing power, delivering better results at a lower cost.

The new NVIDIA Nemotron Nano 2 and Llama Nemotron Super 1.5 models offer the highest accuracy in their size categories for scientific reasoning, math, coding, tool-calling, instruction-following, and chat. These models empower AI agents to think more deeply and work more efficiently—exploring broader options, accelerating research, and producing smarter results within DeFined time limits. Nemotron Nano 2 provides up to 6x higher token generation compared with other leading models of its size. Llama Nemotron Super 1.5 achieves leading performance and the highest reasoning accuracy in its class, enabling AI agents to reason better, make smarter decisions, and handle complex tasks independently. It is now available in NVFP4 (4-bit floating point), delivering up to 6x higher throughput on NVIDIA B200 GPUs compared with NVIDIA H100 GPUs. The Nemotron model family delivers top reasoning accuracy within the Same timeframe and on the same compute budget, providing the highest accuracy per dollar.

Think of the model as the brain of an AI agent—it provides the core intelligence. However, to make that brain useful for a business, it must be embedded into an agent that underStands specific workflows, industry and business jargon, and operates safely. NVIDIA helps enterprises bridge this gap with leading libraries and AI blueprints for onboarding, customizing, and governing AI agents at scale.

Along with the two new Nemotron models, NVIDIA also announced its first open vision language model (VLM) training dataset—Llama Nemotron VLM dataset v1—containing 3 million samples of optical character reCognition, visual QA, and captioning data that power the previously released Llama 3.1 Nemotron Nano VL 8B model. In addition to reasoning model accuracy, agents also depend on retrieval-augmented generation to fetch the latest and most relevant Information from connected data across disparate sources for informed decision-making. The recently released Llama 3.2 NeMo Retriever embedding model tops three visual document retrieval leaderboards—ViDoRe V1, ViDoRe V2, and MTEB VisualDocumentRetrieval—for boosting agentic system accuracy. Using these reasoning and information retrieval models, a deep research agent built with the AI-Q NVIDIA Blueprint currently ranks No. 1 for open and portable agents on the DeepResearch Bench. NVIDIA NeMo and NVIDIA NIM microservices support the entire AI agent lifecycle, from development and deployment to monitoring and optimization of agentic systems.

Cosmos Reason: A Breakthrough in physical AI

VLMs marked a breakthrough for computer vision and robotics, enabling machines to identify objects and patterns. However, non-reasoning VLMs lack the ability to understand and interact with the real world—they cannot handle ambiguity, novel experiences, or solve complex multistep tasks. Cosmos Reason is a new reasoning vision language model for physical AI applications that excels in understanding how the real world works, using structured reasoning to grasp concepts like physics, object permanence, and space-time alignment. The VLM topped the Physical Reasoning Leaderboard on Hugging Face.

Cosmos Reason is a new open, customizable, 7-billion-parameter reasoning VLM for physical AI and Robotics. It allows robots and vision AI agents to reason like humans, using prior knowledge, physics understanding, and common sense to comprehend and act in the physical world. Cosmos Reason enables advanced capabilities across robotics and physical AI applications, including training data critiquing and captioning, robot decision-making, and video analytics AI agents. It is purpose-built to serve as the reasoning backbone for a robot vision language action (VLA) model, or to critique and caption training data for robotics and autonomous vehicles, and to equip runtime visual AI agents with spatial-temporal understanding and reasoning of physical Operations in settings like factories or cities.

Cosmos Reason can help automate the curation and annotation of large, diverse training datasets, accelerating the development of high-accuracy AI models. It can also function as a sophisticated reasoning engine for robot planning, paRSIng complex instructions into actionable steps for VLA Models, even in unfamiliar environments. Furthermore, it powers video analytics AI agents built on the NVIDIA Blueprint for video search and summarization (VSS), enabled by the NVIDIA Metropolis platform, extracting valuable insights from massive volumes of stored or live video data. These visually perceptive and interactive AI agents can help streamline operations in factories, warehouses, retail stores, airports, traffic interSECtions, and more by spotting anomalies. NVIDIA’s robotics research team uses Cosmos Reason for data filtration and curation, and as the “System 2” reasoning VLM behind VLA models such as the next versions of NVIDIA Isaac GR00T NX.

Industry Adoption Across Enterprises

Diverse enterprises and consulting leaders are adopting NVIDIA’s latest reasoning models. Leaders spanning cybersecurity to telecommunications are working with Nemotron to build Enterprise AI Agents. Zoom plans to harness Nemotron reasoning models with Zoom AI Companion to make decisions and manage multistep tasks, taking action for users across Zoom Meetings, Zoom Chat, and Zoom documents. CrowdStrike is testing Nemotron models to enable its Charlotte AI agents to write queries on the CrowdStrike Falcon platform. AMDocs is using NVIDIA Nemotron models in its amAIz Suite to drive AI agents handling complex, multistep automation across care, sales, network, and customer support. EY is adopting Nemotron Nano 2, given its high throughput, to support Agentic AI in large organizations for tax, risk management, and finance use cases. NetApp is currently testing Nemotron reasoning models so that AI agents can search and analyze business data. DataRobot is working with Nemotron models for its Agent Workforce Platform for end-to-end agent lifecycle management. Tabnine is working with Nemotron models for suggesting and automating coding tasks on behalf of developers. Additional agentic AI software developers integrating Nemotron models into their platforms include Automation Anywhere, CrewAI, and Dataiku.

Leading companies across transportation, safety, and AI intelligence are using Cosmos Reason to advance Autonomous Driving, video analytics, and road and workplace safety. Uber is exploring Cosmos Reason to analyze autonomous vehicle behavior and is post-training the model to summarize visual data and analyze scenarios like pedestrians walking across highways to perform quality analysis and inform autonomous driving behavior. Cosmos Reason can also serve as the brain of autonomous vehicles, letting robots interpret environments and, given complex commands, break them down into tasks and execute them using common sense, even in unfamiliar settings. Centific is testing Cosmos Reason to enhance its AI-powered video intelligence platform, enabling the processing of complex video data into actionable insights to reduce false positives and improve decision-making efficiency. VAST is advancing real-time urban intelligence using NVIDIA Cosmos Reason with its AI operating system to process massive video streams at scale. With the VSS Blueprint, VAST can build agents that identify incidents and trigger responses, turning video streams and Metadata into actionable, proactive public safety tools. Ambient.ai is working with Cosmos Reason’s temporal, physics-aware reasoning to enable automated detection of missing Personal protection equipment and monitoring of hazardous conditions, helping enhance environmental health and safety across construction, manufacturing, logistics, and other industrial settings. Magna is developing with Cosmos Reason as part of its City Delivery Platform—a fully autonomous, low-cost solution for instant delivery—to help vehicles adapt more quickly to new cities, with the model adding world understanding to the vehicles’ long-term trajectory planning.

Availability

These models are expected to be available as NVIDIA NIM microservices for secure, reliable deployment on any NVIDIA-accelerated infrastructure for maximum privacy and control. They are planned to be available soon through Amazon Bedrock and Amazon SageMaker AI for Nemotron models, as well as through Azure AI Foundry, Oracle Data Science Platform, and Google Vertex AI. Cosmos Reason can be tried on build.nvidia.com or downloaded from Hugging Face or GitHub. Nemotron Nano 2 and Llama Nemotron Super 1.5 (NVFP4) will be available soon for download. The Llama Nemotron VLM Dataset v1 can be downloaded from Hugging Face. For step-by-step workflows, technical recipes, and concrete examples for building, adapting, and deploying Cosmos WFMs, visit the Cosmos Cookbook or join the commUnity to learn with peers.