FourCastNet predicts global atmospheric dynamics of various weather / climate variables.
Earth-2 FourCastNet NIM
FourCastNet (FCN) predicts accurate short to medium-range global predictions at a time-step size of 6 hours with predictive stability for over a year of simulated time (1,460 steps), while retaining physically plausible dynamics. The multivariant forecast includes surface and atmospheric variables such as wind speed, temperature and pressure at various vertical levels.
The latest version (v2.0.0) uses FourCastNet 3, which advances global weather modeling by implementing a scalable, geometric machine learning approach to probabilistic ensemble forecasting. FourCastNet 3 delivers forecasting accuracy that surpasses leading conventional ensemble models and rivals the best diffusion-based methods, while producing forecasts 8 to 60 times faster than these approaches.
The container components are ready for commercial/non-commercial use.
License/Terms of Use
The NIM container is governed by the NVIDIA AI Product Agreement; and the use of this model is governed by the NVIDIA AI Foundation Models Community License.
Deployment Geography:
Global
Release Date:
1/20/2026 via NGC
Program Classes
The following models are housed within the Earth-2 FourCastNet NIM Container (depending on container version).
| Model Name & Link | Use Case | How to Pull the Model |
|---|---|---|
| FourCastNet 3 (v2.0.0+) | Ensemble Global Weather Forecasting | Automatically downloaded in container version v2.0.0+ |
| FourCastNet 2 (SFNO) (v1.0.0 - v1.1.0) | Deterministic Global Weather Forecasting | Automatically downloaded in container version v1.0.0 - v1.1.0 |
Deployment Details
NVIDIA NIM, part of NVIDIA AI Enterprise, is a set of easy-to-use microservices designed to accelerate deployment of generative AI across cloud, data center, and workstations.
Benefits of self-hosted NIMs:
- Deploy anywhere and maintain control of generative AI applications and data
- Streamline AI application development with industry standard APIs and tools tailored for enterprise environments
- Prebuilt containers for the latest generative AI models, offering a diverse range of options and flexibility right out of the gate
- Industry-leading latency and throughput for cost-effective scaling
- Support for custom models out of the box so models can be trained on domain specific data
- Enterprise-grade software with dedicated feature branches, rigorous validation processes, and robust support structures
Getting Started with FourCastNet NIM
Please visit the Earth-2 NIM Documentation on how to get started.
Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA’s hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.
Changelog
Version v2.0.0
New Model: FourCastNet 3
This version upgrades from FourCastNet 2 (SFNO) to FourCastNet 3, bringing significant enhancements:
- Probabilistic Ensemble Forecasting: Accurately models spatially correlated probabilistic nature with excellent calibration and realistic spectra at extended lead times (up to 60 days)
- Superior Performance: Delivers forecasting accuracy that surpasses leading conventional ensemble models and rivals the best diffusion-based methods, while being 8 to 60 times faster
- Extended Forecasting Range: Maintains stable spectra and realistic dynamics from medium-range to subseasonal timescales
- Rapid Inference: Produces a 60-day global forecast at 0.25°, 6-hourly resolution in under 4 minutes on a single GPU
Key Capabilities:
- 72 surface and pressure-level variables
- 0.25 degree resolution (721 x 1440 grid)
- Rollout stability at subseasonal timescales
- Trained on ERA5 data (1980-2019)
For more details, see the FourCastNet 3 paper.
Version v1.1.0 and v1.0.0
These versions used FourCastNet V2 (SFNO) for global weather forecasting.
Security Vulnerabilities in Open Source Packages
Please review the Security Scanning tab on NGC to view the latest security scan results.
For certain open-source vulnerabilities listed in the scan results, NVIDIA provides a response in the form of a Vulnerability Exploitability eXchange (VEX) document. The VEX information can be reviewed and downloaded from the Security Scanning tab.
Ethical Considerations:
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 internal developer team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns here.
Get Help
Enterprise Support
Get access to knowledge base articles and support cases or submit a ticket.
NVIDIA AI Enterprise Documentation
Visit the NVIDIA AI Enterprise Documentation Hub for release documentation, deployment guides and more.