RoboFacetRoboFacet

NVIDIA Cosmos is a world foundation model platform for developing physical AI such as robots and autonomous vehicles.

Documentation in progress

The product identity and source are confirmed, but some specifications or interface details do not yet have citable manufacturer evidence. Missing evidence is not treated as a negative claim.

Product information

Information below comes from the product website or public documentation. Pricing and availability may change.

Category
Robot Development
Maturityi
Announced
Deploymenti
Not specified
Buying / access
Buy or sign up directly
Source
Official websiteChecked Oct 10, 2026

Reality check

A demo shows what may be possible. These points help judge whether the product fits day-to-day use.

Works well when
  • Simulation, control, perception and repeatable experiments
  • Teams that can manage dependencies and hardware interfaces
  • Prototyping before committing to a physical deployment
What to check
  • Supported hardware, GPU requirements and version compatibility
  • Documentation, examples and community response
  • How simulation results transfer to the real robot
Not a magic solution
  • A guarantee that simulation equals a production cell
  • A replacement for safety validation on hardware
  • A low-maintenance toolchain without version ownership

About

NVIDIA Cosmos is a platform from NVIDIA comprising generative world foundation models, tokenizers, guardrails and an accelerated video processing pipeline for physical AI development. It generates physics-based synthetic data from inputs such as text, image, video, robot sensor or motion data, and supports video search, model development and evaluation, and simulation. The models are available under an open model license on Hugging Face and the NVIDIA NGC catalog, with adoption cited among robotics and automotive companies including 1X, Agility, Figure AI, Waabi and XPENG.

Scenes

Product details

Manufacturer / team
NVIDIA
Product typei
Platform

The cost iceberg

The purchase price is only one part of an automation project.

GPU, simulation assets and hardware adapters
Dependencies, datasets and test infrastructure
Version pinning, documentation and support
Sim-to-real gaps and safety validation