earth2studio and earth2mip
Earth2studio is a general-purpose framework for building and deploying weather/climate AI models, while earth2mip is a specialized research tool for comparing multiple such models against each other, making them complementary tools used together in model development and evaluation workflows.
About earth2studio
NVIDIA/earth2studio
Open-source deep-learning framework for exploring, building and deploying AI weather/climate workflows.
Provides a unified interface to multiple state-of-the-art AI weather models (GraphCast, Pangu, Aurora, AIFS, StormCast) with pluggable data sources (GFS, IFS, satellite observations) and output backends, enabling composition of complex multi-model pipelines. Built as a modular inference toolkit supporting prognostic models for time-series forecasting, diagnostic models for derived quantities, and data assimilation workflows, all sharing a consistent API across different underlying frameworks.
About earth2mip
NVIDIA/earth2mip
Earth-2 Model Intercomparison Project (MIP) is a python framework that enables climate researchers and scientists to inter-compare AI models for weather and climate.
Scores updated daily from GitHub, PyPI, and npm data. How scores work