Berlin, Germany
Hourly precipitation
Rain bands · next 24 hours
Global forecasts, every hour of the day
WeatherNext 3 AI Weather Forecasting
Our most advanced global weather AI model yet. Explore faster, higher-resolution forecasts built from satellite data for smarter, better-informed decisions.
Forecasts every hour
Direct satellite inputs
5 km local variables
Built for real decisions
Explore with WeatherNext 3
Explore Weather with WeatherNext 3
Ask about a location, select the variables that matter, and move into a weather-focused AI experience for planning and analysis.
AI ModelWeatherNext 3
HourlySatelliteVariables
Forecast cadence
Time horizon
LIVE FORECAST LAYERGlobal surface variables
Hourly · high resolution · satellite-driven
FORECAST WORKFLOW
Live, Local, Actionable Data
Combine a location, weather variable, time horizon, and decision context to focus your exploration.
View guideForecast starting points
WeatherNext 3 Forecast Examples
Explore the kinds of hourly, local, and industry-focused weather questions WeatherNext 3 is designed to support.
Tokyo, Japan
Surface temperature
5 km resolution · hourly
North Sea
Wind generation
Wind speed · energy planning
Rocky Mountains
Snow conditions
Fast-changing conditions
California, USA
Solar production
Radiation and cloud cover
Atlantic Ocean
Tropical systems
Global forecast layers
Live, local, actionable
Key Capabilities of WeatherNext 3
Better reach, speed, and detail for decisions that depend on the weather.
Forecast Every Hour
Track fast-changing rain and snow with forecasts generated every hour of every day.
Direct Satellite Data
WeatherNext 3 draws directly from raw satellite imagery to expand reach and detail.
Higher Local Resolution
Predict targeted temperature and humidity at 5 km and other surface variables at 10 km.
Global Surface Variables
Explore precipitation, temperature, humidity, wind, radiation, and cloud cover.
Industry Applications
Support renewable-energy planning and weather-sensitive operations with relevant variables.
Operational Access
Use high-resolution forecasts without setting up and operating a forecasting model yourself.
Four-step workflow
How to Explore WeatherNext 3
Move from a place and a weather question to focused forecast intelligence.
Choose a Location
Enter a city, region, or operational area you want to understand.
Select Variables
Choose precipitation, temperature, humidity, wind, radiation, or cloud cover.
Describe Your Decision
Add the question, timing, and context that matter to your planning.
Explore the Forecast
Generate a weather-focused conversation and examine the conditions that matter.
EXAMPLE QUESTION
How could hourly wind and cloud-cover forecasts affect renewable energy operations in this region tomorrow?
Choose the right path
Which Forecast Workflow Should You Use?
Start with the information you have and the decision you need to make.
WorkflowWhat you provideGood for
Local ForecastA place, variables, and time rangeHourly rain, snow, temperature, humidity, and wind
Operational PlanningA location and business decisionEnergy, logistics, agriculture, and weather-sensitive operations
Forecast AnalysisVariables, region, and comparison goalResearch and historical or operational data exploration
Real-world decisions
WeatherNext 3 Use Cases
Bring frequent, detailed weather information into the decisions people and organizations make every day.
Renewable Energy
Plan wind and solar operations using wind, radiation, and cloud-cover forecasts.
Explore forecastLocal Planning
Understand hourly temperature, humidity, rain, and snow conditions for a specific place.
Explore forecastLogistics
Bring weather context into route, staffing, and time-sensitive operational decisions.
Explore forecastAgriculture
Follow local precipitation, heat, humidity, and wind variables that affect field work.
Explore forecastExtreme Weather
Explore changing forecast layers and tropical cyclone context alongside official guidance.
Explore forecastResearch
Compare high-resolution operational and historical forecasts for analysis and experimentation.
Explore forecastWEATHERNEXT IN ACTION
Supporting better hurricane prediction
See how WeatherNext helped the National Hurricane Center better predict Hurricane Melissa’s historic landfall in Jamaica.
Explore weather intelligenceOperational access
Choose a WeatherNext 3 Access Path
Access high-resolution operational and historical forecasts without setting up the model yourself.
Research Data
Explore published WeatherNext research datasets and compare forecast behavior over time.
Forecast Layers
Inspect precipitation, wind, temperature, humidity, radiation, and cloud-cover layers.
Weather Lab
Use an experimental environment to explore tropical systems and compare model outputs.
Local Scenarios
Frame a place, time horizon, and weather question for a focused exploration.
Research-focused exploration paths Operational forecasts Historical forecasts High-resolution variables Decision support research
Frequently asked questions
WeatherNext 3 FAQ
Answers about WeatherNext 3 forecasts, resolution, variables, access, and applications.
What is WeatherNext 3?+
WeatherNext 3 is Google DeepMind’s most advanced global weather AI model. It generates forecasts every hour and delivers high-resolution local weather variables.
How often does WeatherNext 3 generate forecasts?+
It is the first global WeatherNext model designed to generate forecasts every hour of the day, every day of the year.
What data does the model use?+
WeatherNext 3 directly leverages raw satellite imagery, enabling frequent global forecasts and detailed local outputs.
What resolution does WeatherNext 3 provide?+
It predicts weather-station-targeted temperature and humidity at 5 km resolution and other surface variables such as wind at 10 km.
Which weather variables are supported?+
The published capabilities include precipitation, snow, temperature, humidity, wind, radiation, and cloud cover.
Can WeatherNext 3 support renewable energy?+
Yes. It forecasts variables relevant to wind and solar farms, including wind, radiation, and cloud cover.
What is Weather Lab?+
Weather Lab is an experimental research platform for exploring global forecast layers, tracking tropical cyclones, and comparing AI models with meteorological baselines.
Is this an official Google DeepMind website?+
No. weathernext3.lol is an independent third-party experience and is not operated by or affiliated with Google DeepMind.
Turn Weather Data Into Better Decisions
Start with a place, a weather variable, and the decision you need to make.