For many businesses, weather isn’t an afterthought but a variable that directly affects planning, logistics and revenue. Yet most business applications still rely on weather data from a separate, external provider, or have none at all. Google Maps Platform has an answer to that: the Weather API. And since 3 September 2026, that API has been powered by WeatherNext 3, the latest AI weather model from Google DeepMind and Google Research. As a Google Premier Partner, we’re happy to explain what this means for you.
What is the Google Weather API?
The Weather API is part of Google Maps Platform’s Environment product line, alongside the Air Quality, Pollen and Solar APIs. You provide a coordinate and get back current weather conditions, forecasts and recent history in a developer-friendly format. It’s exactly the same weather data Google itself uses in Google Search, Google Maps and the Gemini app.
In concrete terms, the API offers five types of data:
- Current conditions: temperature, feels-like temperature, humidity, dew point, precipitation (rain, snow, ice), air pressure, UV index, visibility, wind (speed, direction, gusts) and cloud cover.
- Hourly forecast up to 240 hours ahead: ten days of hourly data, for applications where timing matters.
- 10-day daily forecast: including sunrise, sunset and moon phases.
- 24-hour history: what happened hour by hour over the past day, useful for analysis and reporting.
- Weather alerts: real-time alerts for extreme situations such as storms, floods or wildfires.
The data is refreshed every 15 to 30 minutes and continuously monitored for accuracy. That’s a fundamental difference from providers that publish a new run only a few times a day.
What changes with WeatherNext 3?
This is the real news. Traditional weather models solve physics equations on supercomputers and need about six hours to gather all observations before a new calculation starts. WeatherNext 3 works differently: the model learns directly from raw satellite imagery and weather stations, and refreshes its forecast every hour at a resolution of 5 kilometres. By comparison, its predecessor WeatherNext 2 worked at 25 kilometres on a six-hour cycle.
According to Google, this delivers up to 50% more accurate precipitation forecasts for a day or more ahead. The words “up to” matter here: it’s an upper limit, not an average. But the direction is clear. A model that learns every hour from what’s actually happening in the atmosphere will spot an approaching shower over Rotterdam sooner than a model still working with the situation at six o’clock this morning.
For you as a developer or product owner, nothing changes technically. Same API, same endpoints, same integration. The model under the hood has simply got better.
What is the Weather API used for?
Google names four main use cases, and we recognise all of them in the Dutch market:
Operational planning. Think of staff scheduling at a beach bar or theme park, planning roofing work, or deciding when a construction crane can or can’t operate. With a ten-day hourly forecast, you plan on data rather than gut feeling.
Logistics and routing. Combine the Weather API with the Routes API and you can adjust trips dynamically based on icy roads, fog or storms. For fleet operators, that means fewer delays and safer driving.
Travel and leisure. Booking platforms, campsites and outdoor apps show visitors what the weather will be like at their destination, combined with air quality and pollen if you add the other Environment APIs.
Retail and marketing. Demand for ice cream, umbrellas, barbecue meat or winter tyres is directly tied to the weather. With weather data in your inventory and campaign systems, you can steer promotions and restocking based on what’s really going to happen next week. L’Oréal, for example, uses the Environment APIs to tailor skincare advice to the customer’s local weather and air quality.
Weather API or WeatherNext data in BigQuery?
It’s a question we get asked regularly. Google also offers the raw WeatherNext model output through BigQuery, Earth Engine and Cloud Storage. That’s intended for research and heavy grid-level analysis, for example energy companies that want to model wind forecasts at turbine height.
The Weather API is the practical option: a ready-made answer per location, directly usable in your app or website, combined with traditional models where that improves quality. For 95% of business applications, the API is the right starting point.
Try it yourself
Google offers a free Maps Demo Key that lets you prototype without a credit card. There’s also an interactive demo to explore the data, and the developer documentation for the technical details.
How can we help?
As a Google Premier Partner specialising in Location-Based Services, we help organisations turn weather data into a working application. That often starts with a simple question: where does weather play a role in your process, and what would you do differently if you knew it hour by hour, ten days ahead?
From there, we build a proof of concept, connect the Weather API to your existing Google Maps Platform integration and keep an eye on costs. Not sure whether your current Google Maps setup is efficiently configured? Our Quick Scan maps out your API usage and costs.
Want to know more about the Environment APIs? Check out our Environment page or get in touch. We’re happy to help.
