Share:
On this page, we want to introduce you to our models, from which we get forecasts of atmospheric and ocean conditions.
Windy.app has two main types of forecast models: global — with forecasts at each point on the Earth, and regional — with forecasts for particular areas or countries. Regional models pay close attention to the specific geographical features and surface conditions, but they offer a shorter-term forecast.
This is a comparison table and the main features of each model:
Weather models comparison table. Illustration: Valerya Milovanova / Windy.app
Let's understand what the data in the table means:
Now more about each model:
ECMWF weather model, London weather forecast
ECMWF (European Centre for Medium-Range Weather Forecasts) is a global weather model created and operated by the European Centre for Medium-Range Weather Forecasts with headquarters in Reading, Berkshire, United Kingdom, since December 9–10, 1971. The main feature of the model is that it is one of the two leading global weather models on par with GFS from the United States. It is also considered the best for precipitation and cloudiness.
Learn more about ECMWF weather model in the Windy.app blog
GFS27 weather model, New York weather forecast
GFS27 (Global Forecast System) is a global weather model by the United States' National Weather Service.
Learn more about GFS weather model in the Windy.app blog
GFS+ weather model, Los Angeles weather forecast
GFS+ (Global Forecast System +) is Windy.app's modification of a global weather model by the United States' National Weather Service. While the GFS27 interpolates data for every dot within a 27 km x 27 km square, GFS+ always shows only the maximum result in each square.
Learn more about GFS weather model in the Windy.app blog
ICON13 weather model, Rio de Janeiro weather forecast
ICON13 (Icosahedral Nonhydrostatic) is a global weather model by the Deutscher Wetterdienst (German National Meteorological Service).
Learn more about ICON weather model in the Windy.app blog
EXP3 weather model, Bellagio weather forecast
EXP3 is a unique, high-resolution wind and gusts forecast model developed by the Windy.app team. Unlike traditional low-resolution global forecasts, EXP3 uses AI to provide detailed, terrain-aware predictions of wind flow along coastlines. The model is optimized for coastal zones and complex terrain, and covers coastal areas around the world.
Learn more about EXP3 weather model in the Windy.app blog
ICON7 weather model, Munich weather forecast
ICON7 (Icosahedral Nonhydrostatic) is a regional European weather model by the Deutscher Wetterdienst (German National Meteorological Service).
Learn more about ICON weather model in the Windy.app blog
ICON-D2 weather model, Vienna weather forecast. Jacek Dylag / Unsplash
ICON-D2 (Icosahedral Nonhydrostatic) is a regional hyper-local weather model for Central Europe by the DWD (Deutscher Wetterdienst), German National Meteorological Service. Thanks to its high resolution, it takes into account the terrain and also predicts severe weather better.
Learn more about ICON weather model in the Windy.app blog
ICON-2I weather model, Vieste weather forecast
ICON-2I is a high-resolution regional version of the ICON weather model for Italy and the surrounding seas, including the Adriatic Sea. It fills the gap in local weather models for the Mediterranean, especially for the Adriatic — one of the most popular sailing areas, where conditions are complex and require close attention to the weather.
Learn more about ICON weather model in the Windy.app blog
NAM weather model, Chicago weather forecast
NAM (North American Mesoscale Forecast System) is a regional North American weather model by the National Centers for Environmental Prediction (NCEP) of the National Oceanic and Atmospheric Administration of the US (NOAA).
Learn more about NAM weather model in the Windy.app blog
HRRR weather model, San Francisco weather forecast
HRRR (High Resolution Rapid Refresh) is a regional North American weather model by the National Oceanic and Atmospheric Administration of the US (NOAA). It provides a great forecast for the continental US, Canada, and Mexico based on radar data — the unique advantage of HRRR. Radar data is assimilated every 15 min over a 1 h period. It gives short-range weather forecasts with pretty good accuracy.
Learn more about HRRR weather forecast model in the Windy.app blog
RRFS weather model, Ala Moana Beach weather forecast
RRFS (Rapid Refresh Forecast System) is a regional weather model by the National Oceanic and Atmospheric Administration of the US (NOAA) and one of the best models for the United States. It gives a forecast for up to 84 hours ahead — much longer than the 36 hours of HRRR. NOAA plans to make RRFS its main short-range forecasting system, replacing HRRR and NAM over time. RRFS covers not only the continental US but also Alaska, Hawaii, and Puerto Rico.
HRDPS (High Resolution Deterministic Prediction System) is a highly detailed Canadian weather model. The high resolution allows the model to take into account complex terrain and topography. It can reproduce breezes and local circulations, providing accurate mean wind speed, direction, gusts, air temperature, precipitation, cloudiness, atmospheric pressure at sea level, and relative humidity.
AROME weather model, Paris weather forecast
AROME is a highly detailed regional weather model by Météo-France (French National Meteorological Service). It is calculated for France and the surrounding territories and takes into account terrain and local processes.
Learn more about AROME weather model in the Windy.app blog
AROME Tropics is a version of the AROME model designed specifically for the French overseas territories. It originally consists of five separate regional segments that we combined together: Réunion–Mayotte, French Guiana, the Antilles, New Caledonia, and French Polynesia.
The high resolution of this model significantly improves forecast accuracy in coastal areas.
ACCESS-C (Australian Community Climate and Earth-System Simulator – Convective) is a regional Australian weather model developed by the Bureau of Meteorology. It covers Australia’s capital city regions and their surroundings. The model excels at forecasting wind, precipitation, and extreme weather events and is especially well known for its high accuracy in coastal areas.
UKV2 is a Met Office weather model for the UK, Ireland, and parts of Western Europe. It’s detailed enough to show breezes and local circulations in the wind forecast, and its accuracy makes it perfect for short-term planning.
MFWAM weather model, Maui, Hawaii, USA, weather forecast
MFWAM (Météo-France WAve Model) is a third-generation numerical wave forecast model provided by Météo-France, the French National Meteorological Service, which uses it as the basis of its global wave system.
Learn more about MFWAM weather model in the Windy.app blog
GFS-Wave is based on WAVEWATCH III®, an advanced third-generation wave model initially developed at NOAA/NCEP. It builds upon earlier iterations, namely WAVEWATCH and WAVEWATCH II, but distinguishes itself through significant enhancements in governing equations, model structure, numerical methods, and physical parameterizations. This version introduces specialized source term options tailored for extremely shallow water scenarios, including the surf zone, alongside wetting and drying functionalities for grid points. While surf-zone physics in this version is basic, it extends the model's utility to arbitrary shallow water environments.
MyOcean weather model, Palermo, Sicily, Italy, weather forecast
The MyOcean weather model is the world's main model with currents, tides, and sea temperature forecasts. These are typically updated once a day and cover oceans, seas, and coastal areas, but not rivers and lakes, even big ones. The model is part of the Copernicus Marine Service (CMS), which, in turn, is run by the Copernicus Programme, the European Union's Earth Observation Programme, looking at our planet and its environment for the ultimate benefit of all European and global citizens.
Learn more about MyOcean weather model in the Windy.app blog
Get the list of all the world's major weather models, including the ones that are not used in the Windy.app.
Updated on October 14, 2026
Cover photo: Markusspiske / Unsplash
