Alternative to OpenWeather One Call: Keep One Call Meaning One Call

OpenWeather is moving One Call 3.0 users to One Call 4.0, a new API with a separate subscription. The data is now split across six endpoints and long forecasts are paginated, so a request that used to be one billed call now takes six or more.

08. 10. 2026
Rain-covered window overlooking Big Ben and the Houses of Parliament, illustrating weather forecast data

If you want one call to still mean one call, the Meteosource weather API is one of the easiest ways to replace OpenWeather One Call. Our forecasts blend 40+ global weather models with machine learning for higher accuracy, backed by 20+ years of forecasting experience. The minute-by-minute precipitation forecast updates every 10 minutes and the short-term forecast every hour, so your data stays fresh.

What changes in One Call 4.0

In One Call 3.0, one request returned current weather, minutely, hourly and daily forecasts and alerts. In One Call 4.0, the same data needs at least six billed calls: current (1), minutely (1), 48 hours of hourly data at 20 records per page (3) and daily (1), plus one extra call for each weather alert. OpenWeather's own migration guide confirms that one 3.0 call can become several billed calls in 4.0.

At £0.0012 per call above the free tier, the lower price per call doesn't make up for six times as many calls.

Comparison of features

Feature OpenWeather One Call 4.0 Meteosource API
Current weather conditions Yes Yes
Minute-by-minute forecast 60 minutes 60 minutes
Hourly forecast 2 days 3 days (Startup), 7 days (Standard)
Daily forecast Up to 1.5 years (long-range estimates) 10 days (Startup), 30 days (Standard), model-based forecast
Severe weather alerts Separate call per alert Included in the response
Solar irradiance (GHI) Not available in any self-service plan Included (Standard)
Renewables power forecast (solar and wind) Not available in any self-service plan Renewables plan, £15/location/month
Calls per forecast 2 to 6, depending on the data you need 1
Free tier 1,000 calls/day = about 166 complete forecasts 400 calls/day = 400 complete forecasts, no credit card
Pricing £1.20 per 1,000 calls above free tier £7/month for 10,000 calls/day, then £1 per 1,000 calls
Cost per 1,000 forecasts £2.40 to £7.20 As low as 2p (Startup plan at full use)

A note on the daily forecast: no weather model forecasts day-by-day weather 1.5 years ahead, so daily values that far out are long-range estimates, not a weather forecast.

How we calculated it: One Call 4.0 bills each type of data separately. If you only need current weather and a daily forecast, that's 2 calls, so 1,000 forecasts = 2,000 calls × £0.0012 = £2.40. For the full set One Call 3.0 returned in one call (current, minutely, 48 hours of hourly data and daily), it's 6 calls: current (1), minutely (1), hourly at 20 records per page (3) and daily (1). That makes 1,000 forecasts = 6,000 calls × £0.0012 = £7.20. Weather alerts add one call each. With Meteosource, a forecast is one call whatever data you need. The Startup plan is £7/month for 10,000 calls a day, which is 300,000 calls a month, or about 2p per 1,000 if you use the full quota. Above the limit it's £1 per 1,000 calls. All prices are without VAT.

What you'd actually pay

Here's what two typical projects cost per month on each API.

Project OpenWeather One Call 4.0 Meteosource
Weather dashboard: 50 locations, current + daily forecast, refreshed every 30 minutes (2,400 forecasts/day) About £137/month £7/month (Startup)
Weather app: 100 cities, full forecast, refreshed every 10 minutes (14,400 forecasts/day) About £3,074/month £22/month (Standard)

Dashboard on OpenWeather: 2 calls per forecast = 4,800 calls a day, minus 1,000 free = 3,800 × £0.0012 = £4.56 a day. Weather app: 6 calls per forecast = 86,400 calls a day, minus 1,000 free = 85,400 × £0.0012 = £102.48 a day. Months are 30 days, weather alert calls are not included and all prices are without VAT.

With Meteosource, a single API call returns current weather plus minute-by-minute, hourly and daily forecasts for any location in the world. Compare all plans on our pricing page.

More accurate data, not just fewer calls

Switching APIs to save money only makes sense if the data holds up. With Meteosource, it does better than that. Most weather services build their forecasts on a single global model and then apply a regional model to it, so whatever that one model gets wrong goes straight into your app. Meteosource builds every forecast in four layers instead:

  1. Machine-learning blend of global models. Our algorithms combine the forecasts of all available global models with years of historical weather data and past forecasts. They learn each model's errors and biases in different weather situations and locations, then produce a single forecast of the most probable scenario, computed four times a day.
  2. High-resolution regional models. On top of the global blend, our regional models add resolution and account for local conditions and geography. A meteorologist on duty can adjust some of the inputs when the situation calls for it.
  3. AI corrections for cities. For places with weather station measurements, usually towns, cities and more populated areas, an extra AI layer adjusts the forecast for local weather patterns.
  4. Real-time nowcasting. Between model runs, measurements from weather stations, radar and other sources keep updating the forecast. That's what makes the minute-by-minute precipitation forecast possible, refreshed every 10 minutes.

The result is a forecast that is more accurate than any single model it is built from, which is why customers who benchmarked several providers tell us Meteosource "outperforms all of them in terms of accuracy". There are years of experience and research behind the API. Read more about how our forecasts are made.

Easy migration from OpenWeather in 4 simple steps

  1. Check our subscription plans and choose either the Free plan or one of our paid plans.
  2. Sign up for free.
  3. Get your API key and follow our documentation or use our client libraries for Python, JavaScript, PHP and C++.
  4. Call all weather data in one simple request.

Just like One Call, Meteosource works with GPS coordinates, and you can also use a place ID. You can set language, time zone and units to get exactly the response you need.

Parameters OpenWeather vs Meteosource

OpenWeather One Call Meteosource API Description
dt date The timestamp of the data point
weather.description summary A human-readable text summary
weather.icon icon A machine-readable weather icon
pop probability.precipitation The probability of precipitation (OpenWeather: 0 to 1, Meteosource: %)
temp temperature Temperature 2 meters above the surface
feels_like feels_like The apparent ("feels like") temperature
dew_point dew_point Dew point
humidity humidity Humidity in percentage
pressure pressure Atmospheric pressure at mean sea level
wind_speed wind.speed Wind speed 10 metres above the ground
wind_gust wind.gusts Wind gust speed
clouds cloud_cover (current), cloud_cover.total (hourly, daily) Cloud cover in percentage
uvi uv_index The UV index
visibility visibility Visibility
  and more variables  

Try Meteosource for free

Sign up for free and get 400 calls a day, no credit card needed. Want to test the paid features first? The Standard plan comes with a free 10-day trial. Need a custom solution or higher volumes? Contact us.

Do you like this article?
Please share it with your friends