Introduction to WeatherNext 3

WeatherNext 3 is an hourly, high-resolution AI weather forecasting experience. This documentation explains what the model produces, how to explore a forecast here, and how plans and forecast credits work.

If you want the background first, read What is WeatherNext 3? and then come back for the operational detail. This site is an independent third-party experience built around WeatherNext 3 capabilities.

What WeatherNext 3 produces

WeatherNext 3 is built to generate forecasts every hour of the day, every day of the year, and to expose local weather variables at a resolution that is useful for real decisions.

Variable groupWhat you getTypical questions
PrecipitationRain and snow timing, intensity, and movementWhen does the rain band arrive?
Surface temperatureTargeted temperature at 5 kmHow warm will the afternoon get?
HumidityLocal moisture levels read alongside temperatureWill the air dry out overnight?
WindSpeed and direction for energy and logistics planningIs the wind window usable?
Radiation and cloud coverSolar-relevant inputs for production planningHow much sunlight reaches the site?
Tropical systemsGlobal forecast layers for large-scale eventsWhere is the system tracking?

Temperature and humidity are resolved at 5 km, and other surface variables are provided at 10 km. Forecasts are refreshed every hour, so fast-changing conditions such as convective rain, snow bands, and shifting wind are visible without waiting for the next daily model cycle.

Getting started

  1. Choose a location. Enter a city, region, or operational area. Coordinates and place names are both accepted.
  2. Select variables. Pick the variables that match your decision, such as precipitation, temperature, humidity, wind, radiation, or cloud cover.
  3. Describe your decision. Add the question, the time horizon, and any operational context. The more specific the question, the more focused the answer.
  4. Explore the forecast. Generate a weather-focused conversation and work through the conditions that matter to you.

The full walkthrough, including worked examples for energy and field operations, is in How to use WeatherNext 3.

Reading a forecast responsibly

  • Treat every forecast as a probability distribution, not a promise. Hourly updates help you see how the distribution shifts over time.
  • Compare the hourly progression instead of a single timestamp. A rain band that moves by two hours still changes the decision.
  • Use local variables together. Temperature alone rarely explains road ice, crop stress, or turbine icing; humidity, wind, and radiation fill in the rest.
  • For safety-critical decisions, always keep official national weather service warnings in the loop.

Plans and forecast credits

Every plan includes forecast credits that renew each month. Credits are consumed as you generate forecasts and conversations, so a plan maps directly to how much exploration you do.

PlanMonthly creditsBuilt for
StarterEntry levelExploring global forecasts and light analysis
ProHigher monthly allocationRegular operational and planning work
StudioHighest allocationTeams running many locations or scenarios

See pricing for current prices, annual discounts, and what each tier unlocks.

Accessing forecast data programmatically

If you want to move from manual exploration to automated workflows, read WeatherNext 3 API and data access for how teams typically request forecast fields for a location and time range, and what to consider before wiring forecasts into a production system.

Guides

Frequently asked questions

How often are forecasts updated?

WeatherNext 3 is designed around hourly forecast generation, so new output is available throughout the day rather than once per model cycle.

Which variables are available at 5 km?

Targeted temperature and humidity outputs are available at 5 km, while other surface variables such as wind, precipitation, radiation, and cloud cover are provided at 10 km.

Do I need my own model or compute?

No. The point of exploring WeatherNext 3 here is to use high-resolution forecasts without setting up and operating a forecasting model yourself.

Introduction to WeatherNext 3