1. [Home ](/) Metropolis # NVIDIA Metropolis NVIDIA Metropolis is a collection of models, libraries, and blueprints that provides everything you need to build, deploy, and scale [video analytics AI agents](https://www.nvidia.com/en-us/use-cases/video-analytics-ai-agents/) and applications, from the edge to the cloud. You can now easily transform raw video and sensor data from real-world environments into real-time, actionable insights. This helps your organization understand what’s happening in your physical spaces and respond intelligently, while delivering exceptional scale, throughput, cost-effectiveness, and faster time to production. [Get Started ](https://build.nvidia.com/explore/vision "Get Started")[Join the Forum](https://forums.developer.nvidia.com/c/accelerated-computing/intelligent-video-analytics/13 "Join the Forum") * * * ## How Metropolis Works Metropolis offers a cohesive, end-to-end stack of software building blocks that handle everything from video ingestion to insight generation to advanced agentic AI-powered analytics. These components can be deployed consistently across the whole compute spectrum—on the edge, in on-prem servers, or in the cloud—so the same applications can run close to where data is generated or centrally at scale. ![img-alt-text](https://developer.download.nvidia.com/images/metropolis/gtc26-metropolis-stack-update-r2.png) ### Agentic Video Search at Scale With NVIDIA VSS Blueprint Dive into the details of the new VSS 3.0 capabilities in agentic search, modular design, reference workflows, and more. [Learn More](https://docs.nvidia.com/vss/latest/)[Watch the Video](https://www.youtube.com/watch?v=39_rgXBZmsw) ### Create Vision AI Applications With Generative AI Coding Agents Learn how to generate complete, GPU-accelerated NVIDIA DeepStream video analytics pipelines using simple natural language prompts. **Coming Soon** ### Post-Training Recipes for NVIDIA Cosmos Reason VLM See step-by-step workflows, case studies, and technical recipes for NVIDIA Cosmos™ world foundation models. [Explore Cosmos Cookbook](https://github.com/nvidia-cosmos/cosmos-cookbook)[Read the Prompt Guide](https://nvidia-cosmos.github.io/cosmos-cookbook/core_concepts/prompt_guide/reason_guide.html) ### Solve the Training Data Challenge With the Physical AI Data Factory Learn how to build a training data pipeline with your own video or image data. Then, curate, augment, and evaluate it with Cosmos for vision AI models. [Read the Press Release](https://nvidianews.nvidia.com/news/nvidia-announces-open-physical-ai-data-factory-blueprint-to-accelerate-robotics-vision-ai-agents-and-autonomous-vehicle-development) * * * ## Get Started With Metropolis ### Start using the latest Metropolis vision language models and vision foundation models. ![Build And explore with NVIDIA Rivermax SDK](https://developer.download.nvidia.com/icons/m48-cosmos.svg) ### NVIDIA Cosmos Reason Industry-leading physical AI reasoning vision-language model enables video analytics AI agents to see, understand, and act in the physical world like humans. [Try It Now](https://build.nvidia.com/nvidia/cosmos-reason2-8b) [Use the Prompt Guide](https://nvidia-cosmos.github.io/cosmos-cookbook/core_concepts/prompt_guide/reason_guide.html) ![Build And explore with NVIDIA Rivermax SDK](https://developer.download.nvidia.com/icons/nim.svg) ### Vision AI NIMs Discover GPU-optimized microservices that bring ready-to-use vision and multimodal models through simple APIs. [Try NVIDIA NIM™ APIs](https://build.nvidia.com/explore/vision) ![icon representing neural network](https://developer.download.nvidia.com/icons/m48-neural-network-3.svg) ### Embeddings Turn images, videos, and text into vector representations for physical AI and multimodal understanding using NVIDIA models like Cosmos Embed, C-RADIO, and NV-CLIP. [Explore Cosmos Embed](https://docs.nvidia.com/nim/cosmos-embed1/latest/introduction.html) [Explore C-RADIO](https://huggingface.co/nvidia/C-RADIOv4-H) [Explore NV-CLIP](https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/nvclip?version=2.0.0) ### Post-train your vision AI models with domain-specific data to boost accuracy. ![Build And explore with NVIDIA Rivermax SDK](https://developer.download.nvidia.com/icons/m48-cosmos.svg) ### Cosmos Cookbook Access recipes to post-train Cosmos Reason VLM and Cosmos Embed with supervised fine-tuning and reinforcement learning. [Learn More](https://github.com/nvidia-cosmos/cosmos-cookbook) ![icon representing TAO Toolkit](https://developer.download.nvidia.com/icons/m48-custom-data-256px-blk.svg) ### TAO Toolkit Explore this low-code solution that uses transfer learning to fine-tune and optimize vision AI models with your data. [Learn More](https://developer.nvidia.com/tao-toolkit) ### Start developing vision AI applications with foundational Metropolis frameworks. ![https://developer.downicon representing building agentic ai application with large language model](https://developer.download.nvidia.com/icons/building-agentic-ai-application-with-large-language-model.svg) ### Video Search and Summarization (VSS) Blueprint The VSS blueprint lets you build customizable video analytics AI agents to deliver powerful insights with seamless edge-to-cloud integration. [Try It Now](https://build.nvidia.com/nvidia/video-search-and-summarization) ![icon representing video processing streaming](https://developer.download.nvidia.com/icons/video-processing-streaming.svg) ### NVIDIA DeepStream This is a complete streaming analytics toolkit for AI-based multi-sensor processing, video, audio, and image understanding. [Learn More](https://developer.nvidia.com/deepstream-sdk) ### Generate high-quality synthetic data to safely and efficiently train your AI models. ![Icon representing data files](https://developer.download.nvidia.com/icons/data-files.svg) ### Physical AI Dataset Unblock data bottlenecks with an open-source, validated dataset for training vision AI in industries, cities, robotics, and autonomous systems—now free on Hugging Face. [Explore the NVIDIA Physical AI Dataset](https://huggingface.co/collections/nvidia/physicalai-67c643edbb024053dcbcd6d8) ![Icon representing neural network](https://developer.download.nvidia.com/icons/m48-neural-network-3.svg) ### Physical AI Data Factory Blueprint Build a synthetic data generation pipeline with your own video or image data. Then, curate, augment, and evaluate it with Cosmos open world foundation models (WFMs) to accelerate vision AI model development. [Learn More](https://nvidianews.nvidia.com/news/nvidia-announces-open-physical-ai-data-factory-bluep[%E2%80%A6]botics-vision-ai-agents-and-autonomous-vehicle-development) [Explore Cosmos Cookbook](https://github.com/nvidia-cosmos/cosmos-cookbook) ![Icon representing visual data simulation](https://developer.download.nvidia.com/icons/simulation-2.svg) ### Isaac Sim Enable developers to create realistic synthetic data from complex 3D environments to train vision AI models. [Get Started With Event and Actor Generation on Isaac SIM](https://docs.isaacsim.omniverse.nvidia.com/latest/action_and_event_data_generation/index.html) * * * ## Starter Kits ### Develop Video Analytics AI Agents Build intelligent systems that can see, understand, and interact with the world through computer vision and real-time visual reasoning. - [VSS Blueprint Github](https://github.com/NVIDIA-AI-Blueprints/video-search-and-summarization) - [VSS Documentation](https://docs.nvidia.com/vss/latest/) - [VSS Agentic Search Demo video (new)](https://www.youtube.com/watch?v=39_rgXBZmsw) - [VSS Event Reviewer Demo video](https://www.youtube.com/watch?v=0FWZwMum7bw) - [Deploy VSS on Cloud](https://brev.nvidia.com/launchable/deploy?launchableID=env-2tYIjRXL4eMCbH9Az8mJC5WPAI4) ### Build a Vision Inference Pipeline Develop a streaming pipeline with DeepStream that ingests videos, preprocesses frames, and runs optimized vision AI models. - [DeepStream Developer Guide](https://docs.nvidia.com/metropolis/deepstream/dev-guide/) - [DeepStream inference Builder Github](https://github.com/NVIDIA-AI-IOT/inference_builder?ncid=so-yout-238844) - [DeepStream Inference Builder Video Tutorial](https://www.youtube.com/watch?v=yj11L8rFC30) ### Build Agents for Smart City and Warehouse Operation Explore examples for optimized VSS blueprint configuration with sample data, tailored prompts, and report templates. - [VSS smart city example](https://docs.nvidia.com/vss/3.1.0/smartcity-docs/smartcity-toc.html) - [VSS warehouse operation example](https://docs.nvidia.com/vss/3.1.0/warehouse-docs/warehouse-toc.html) ### Post-Train Vision Language Model Refine a vision-language model on task-specific multimodal data so it better aligns visual understanding with domain-specific concepts and instructions. - Recipe:[Intelligent Transportation Post-Training Cosmos Reason for Intelligent Transportation](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/post_training/reason2/intelligent-transportation/post_training.html) - Recipe: [Post-Training Cosmos Reason to Understand Spatial AI in Warehouses](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/post_training/reason1/spatial-ai-warehouse/post_training.html) - Recipe: [Wafer Map Anomaly Classification With Cosmos Reason](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/post_training/reason1/wafermap_classification/post_training.html) ### Fine-Tune Vision Foundation Models Adapt powerful pre-trained vision backbones with targeted domain data so they specialize on your tasks while retaining their broad visual understanding. - [Get Started With NVIDIA TAO Fine-Tuning Microservices](https://docs.nvidia.com/tao/tao-toolkit/latest/text/tao_toolkit_api/index.html) - Tech Blog :[Build a Real-Time Visual Inspection Pipeline](https://developer.nvidia.com/blog/build-a-real-time-visual-inspection-pipeline-with-nvidia-tao-6-and-nvidia-deepstream-8/) - Tech Blog: [Optimizing Semiconductor Defect Classification With Generative AI and Vision Foundation Models](https://developer.nvidia.com/blog/optimizing-semiconductor-defect-classification-with-generative-ai-and-vision-foundation-models/) ### Generate Synthetic Data Create synthetic images and videos to expand training datasets, reduce collection costs, and improve vision model robustness across diverse scenarios. - Recipe: [Generating Synthetic Data for Multi-View Warehouse Detection and Tracking With Cosmos Transfer](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/inference/transfer1/inference-its-weather-augmentation/inference.html) - Recipe: [CARLA Simulator-to-Real Augmentation for Traffic Anomaly Scenarios With Cosmos Transfer](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/inference/transfer2_5/inference-carla-sdg-augmentation/inference.html) - Recipe: [Style-Guided Video Generation With Cosmos Transfer 2.5](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/inference/transfer2_5/inference-image-prompt/inference.html) - Recipe: [LoRA Post-Training for Sports Video Generation](https://nvidia-cosmos.github.io/cosmos-cookbook/recipes/post_training/predict2_5/sports/post_training.html) * * * ## More Resources ![Decorative image representing Developer Community](https://developer.download.nvidia.com/icons/m48-people-group.svg) ### Explore Developer Forums ![Decorative image representing Developer Newsletter](https://developer.download.nvidia.com/icons/m48-email-settings.svg) ### Sign up for the Developer Newsletter ![Decorative image representing Developer Community](https://developer.download.nvidia.com/icons/m48-developer-1.svg) ### Join the NVIDIA Developer Program * * * ## Ethical AI NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/). Develop, deploy, and scale AI-enabled video analytics applications with NVIDIA Metropolis. [Get Started](https://build.nvidia.com/explore/vision)