Google DeepMind's WeatherNext 3 adds real-time satellite data and hourly forecasts across Google products

Google DeepMind and Google Research released WeatherNext 3 on Thursday, describing it as their most advanced and accurate global weather AI model. The new version incorporates real-time satellite data, refreshes hourly, runs at higher resolution, adds precise precipitation forecasting, and includes variables aimed at clean energy planning. Google said it is now integrated across Search, Gemini, Maps, Google Maps Platform and Cloud, where users and researchers can access the data directly. "This is going to be the first time that some of the core variables feed and power a lot of the Google products," Samier Merchant, a Google senior staff engineer, told TechCrunch. TechCrunch reported that WeatherNext 3 has proven the most accurate among leading contenders tested on Operational WeatherBench, a comparison utility built by the startup Brightband that examines metrics such as temperature, wind speed and humidity. According to that comparison, the model beat other deep-learning forecasters from Google, Microsoft, Nvidia and the European Centre for Medium-Range Weather Forecasts, as well as traditional physics-based forecasts from the U.S. National Weather Service and the ECMWF. Most operational forecasts still come from government supercomputers solving physical equations, an approach that is accurate but expensive and comparatively slow. Learned models such as WeatherNext 3 continue a shift in meteorology toward deep learning, with practical consequences for agriculture, renewable energy operators and everyday planning.