On November 17, Google DeepMind and Google Research collaboratively introduced WeatherNext 2, a cutting-edge AI weather forecasting model. This innovative model boasts a computational speed roughly eight times faster than its predecessor, enabling it to generate forecasts with hourly precision. Within the 0-to-15-day forecast window, 99.9% of the predicted variables—such as temperature, wind speed, and humidity—demonstrate superior accuracy compared to the previous model. Furthermore, WeatherNext 2 has the capability to simulate hundreds of potential weather scenarios from a single initial condition, with each simulation completed in under a minute on a single TPU (Tensor Processing Unit). Presently, the forecast data generated by WeatherNext 2 is accessible on Earth Engine and BigQuery, while an early access program has been initiated on the Vertex AI platform of Google Cloud. Moreover, this state-of-the-art technology has been seamlessly integrated into Google Search, Gemini, Pixel Weather, and the Weather API of the Google Maps Platform, with intentions to extend its application to weather information within Google Maps in the upcoming weeks.