Google WeatherNext 3 is the company’s newest AI-powered weather model, and it promises more accurate forecasts across a range of Google products. The improved training model is being used within Google Search, Google Maps, Gemini and the Pixel Weather app, drawing on data delivered through the Google Maps Platform Weather API.
Weather information has become a fragmented experience for many smartphone owners. On iOS, Apple provides a native weather app offering a 10-day forecast, hourly conditions and both 1-hour and 12-hour radar sweeps for early precipitation warnings. On Android, the situation is less straightforward.
Android weather apps and the Pixel advantage
Google discontinued its Google Weather app earlier in the year, leaving owners of non-Pixel Android handsets to rely on third-party options such as The Weather Channel or AccuWeather. Pixel users, however, retain access to the Pixel Weather app, which presents similar information to Apple’s offering but with a more colourful design. The two platforms do not always agree on the current temperature or other weather-related readings.
The absence of a native weather app on standard Android phones has led some to suggest that Google removed the feature so it could position Pixel Weather as a selling point for its own hardware. The Pixel Weather app draws its data from the same Google Maps Platform Weather API now benefiting from the new model.
What WeatherNext 3 changes
According to independent live evaluations carried out by Brightband, WeatherNext 3 is the most accurate and advanced global weather model produced to date. The model learns from observations made in real time, allowing it to deliver more timely predictions, including localised forecasts. It converts raw weather data into a high-resolution forecast every hour, and the sharper resolution combined with AI training produces more accurate results.
These forecasts are then made available to people using Google products around the world. WeatherNext 3 can present forecasters with a view of the global weather picture that is five times sharper than the previous model, WeatherNext 2. That earlier system trained on data generated by supercomputer-driven physics, which carried a six-hour data lag.
By learning from real-time observations rather than relying solely on delayed physics-based calculations, the new model aims to reduce that gap and improve both the timeliness and precision of the information supplied to users. WeatherNext 3 is now employed on Google Search, with the same underlying data feeding through to Google Maps, Gemini and the Pixel Weather app.
Source
Image: phonearena.com
