Google DeepMind launches WeatherNext 3 model
Analysis based on 9 articles · First reported Sep 03, 2026 · Last updated Sep 07, 2026
The launch strengthens Alphabet Inc.'s position in the AI-driven weather forecasting market, potentially increasing adoption of its cloud and data services among energy and logistics companies. It also intensifies competitive pressure on traditional meteorological agencies and rivals, potentially reshaping the commercial weather data landscape.
Alphabet Inc. — Google DeepMind and Alphabet Inc. — Google Research unveiled WeatherNext 3, an advanced AI weather model that delivers hourly global forecasts at up to 5-kilometer resolution, roughly five times sharper than its predecessor WeatherNext 2. Unlike traditional numerical weather prediction models that rely on supercomputer simulations with a six-hour data lag, WeatherNext 3 ingests real-time geostationary satellite imagery and sparse weather station observations, enabling faster and more accurate predictions. The model shows significant improvements in precipitation forecasting, with up to 60% better Continuous Ranked Probability Score against United States — NASA's IMERG dataset, and up to 50% more accurate precipitation forecasts for day-ahead planning. It also provides renewable-energy-specific outputs such as wind speeds at 100 meters and solar radiation data. WeatherNext 3 is now integrated across Alphabet Inc. products including Search, Maps, Gemini, and the Maps Platform Weather API, with data accessible via BigQuery and Earth Engine. Independent evaluations by Brightband rank it as the most accurate global weather model currently available. The release builds on Alphabet Inc. — Google DeepMind's lineage of AI weather models, including GraphCast, GenCast, and WeatherNext 2, and reflects a broader trend of technology companies transforming the economics of weather forecasting.
Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.
Open Dashboard