AIFS

Produces global medium-range weather forecasts at about 31 km resolution, six-hourly.

Field
Weather & climate
Method
Deep learning
Data it takes
gridded-fields
License
Apache-2.0

AIFS Single 2.0 is ECMWF’s data-driven forecast model. It steps a global atmospheric state forward six hours at a time, in place of the numerical integration a physics-based model performs.

It joins FourCastNet 3 in the catalog as a global forecast model. StormCast covers the other case, a convection-allowing nowcast over the continental United States.

Weights are fetched at run time. A forecast also needs a real initial state, so running one means supplying analysis fields for the starting time.

Upstream project → · Paper →

Catalog entry last checked 2026-09-08. All models →

Run AIFS on your data

The lab runs this on its own compute, in a container, with the inputs and parameters recorded alongside the result. Initial scoping conversations are free.

Contact the Lab