Trendora

NVIDIA Cosmos

Trial

Platforms

An open frontier omnimodel family for building systems that perceive, reason, plan, and act in the physical world.

Why it's here

Placed in Trial: 8 article(s) of evidence from 2 source(s), led by model releases, with 6 in the last 30 days. Confidence 62%.

Evidence (8)

  • 7NVIDIA GenAI·8/11/2026model_release
    NVIDIA spotlights open local AI models and agent tools

    NVIDIA highlighted a wave of open-weight models, local inference tools, and agent-focused releases from partners and open source communities in its Local AI blog series. The update includes new models such as Cosmos 3 Edge, MiniMax-H3, Laguna S 2.1, DeepSeek-V4-Flash, Inkling-Small, and Wan-Animate-2, along with Unsloth Desktop and NVIDIA-optimized checkpoints for running them on local RTX, DGX Spark, Jetson, and DGX Station hardware.

  • 8NVIDIA GenAI·8/6/2026model_release
    NVIDIA details Cosmos 3 open world models for physical AI

    NVIDIA says open world models are becoming a core building block for physical AI, helping teams generate synthetic data, simulate future states, and specialize systems for robots, autonomous vehicles, and vision AI. The company highlights Cosmos 3 and Omniverse libraries as part of an open workflow for training, testing, and validating physical AI before real-world deployment.

  • 7Hugging Face Blog·7/27/2026product_launch
    NVIDIA Launches Cosmos-H-Dreams for Real-Time Surgical Robotics Simulation

    NVIDIA introduced Cosmos-H-Dreams, a real-time, action-conditioned generative simulator for surgical robotics. The system distills Cosmos-H-Surgical-Simulator into a causal student model and runs through FlashDreams on a single RTX PRO 6000 GPU, with demos for dVRK suturing and the Versius controller.

  • 8NVIDIA GenAI·7/22/2026open_source
    NVIDIA Open Sources GPU-Accelerated Medical Physics Simulation Framework

    NVIDIA has open sourced Medical Physics Simulation, a GPU-accelerated framework within Isaac for Healthcare for simulating anatomy-device interactions, sensor inputs, and difficult edge-case scenarios. The framework is designed to help medical robotics teams train, test, and evaluate policies in simulation before moving to hardware-heavy validation.

  • 7Hugging Face Blog·7/20/2026model_release
    NVIDIA launches Cosmos 3 Edge on Hugging Face

    NVIDIA has released Cosmos 3 Edge, a 4-billion-parameter open world model, in the Cosmos 3 repository on Hugging Face. The model is designed for edge devices and supports real-time reasoning, vision understanding, and robot action generation on NVIDIA hardware such as RTX PRO GPUs, DGX systems, GeForce RTX GPUs, and Jetson modules.

  • 7NVIDIA GenAI·7/20/2026product_launch
    NVIDIA Highlights Agentic AI and Physical AI at SIGGRAPH

    At SIGGRAPH, NVIDIA is showcasing advances in graphics, neural rendering, simulation and AI for content creation and virtual worlds. The company also highlighted Model Context Protocol integrations, a Synthetic Video Detector NIM microservice, Cosmos 3 Edge for local physical AI, and NVIDIA Agent Toolkit support for MCP-based workflows.

  • 8NVIDIA GenAI·7/6/2026research
    Open Models Shape ICML AI Research

    ICML 2026 papers show open frontier models and open AI infrastructure becoming central to modern AI research, with NVIDIA GPUs and Nemotron cited widely across accepted work. The research spans robot world models, life sciences, synthetic data generation, and autonomous systems, highlighting growing use of open datasets, open weights, and reproducible training stacks.

  • 8Hugging Face Blog·1/5/2026model_release
    NVIDIA launches Cosmos Reason 2 for physical AI

    NVIDIA has released Cosmos Reason 2, an open reasoning vision-language model aimed at physical AI use cases. The model improves spatio-temporal understanding, supports larger context windows, and adds capabilities such as OCR, 2D/3D localization, bounding boxes, and trajectory data for robots and AI agents. It is positioned for applications including video analytics, data annotation, autonomous vehicle training, and robot planning, with deployment options from edge to cloud in 2B and 8B parameter sizes.