Dark Sky vs OpenWeatherMap: Two Weather API Services Compared

In this post, we’re going to review two popular weather service APIs:

Dark Sky vs OpenWeatherMap.

For each API, we’re going to look at:

• API Endpoints
• API Stats
• Pricing
• Unique Features

Make sure to read this entire API showdown until the very end. That way, you can ensure that you make the right decision when choosing a weather service.

The Weather APIs

Before I get into the review itself, I want to quickly let you know some important info on each API.

Dark Sky

Let’s start with the Dark Sky API.

Dark Sky (also formerly known as Forecast.io) started as a Kickstarter campaign that is now a popular weather service with successful mobile apps and boasts that they are the most accurate source of hyperlocal weather information: with down-to-the-minute forecasts for your exact location, you’ll never get caught in the rain again. (It’s like magic!)

Connect to the Dark Sky API
Some key features of the API that make it a top-tier weather API include:

• Forecasts and current conditions
• Global coverage
• Historical data
• Severe weather alerts

OpenWeatherMap

On the other side, we have the OpenWeatherMap API.

OpenWeatherMap is arguably the most popular weather service out there with multiple API services, including:

• Current weather data
• Hourly forecast
• 16 day / daily forecast
• Climate forecast for 30 days
• Bulk downloading
• 5 day / 3-hour forecast data
• and much more

Connect to the OpenWeatherMap API

Which API is best for Real-Time Weather Data?

First, we decided to see which API was better for getting accurate real-time data.

Let’s see who comes out on top.

Dark Sky

We’ll start with the Dark Sky API.

To get the current weather conditions, Dark Sky offers a GET endpoint called Forecast, which returns:

the current weather conditions, a minute-by-minute forecast for the next hour (where available), an hour-by-hour forecast for the next 48 hours, and a day-by-day forecast for the next week.

Connect to the Dark Sky APITo begin, we’ll need to enter the lat long coordinates (required parameters) of our target location.

Optionally, we can specify which measurement of units we’d like the stats to be returned in. See more details here.

Let’s see how San Francisco is doing.

When we test the endpoint with these coordinates, we get this JSON response:

Not only does the response provide current weather stats such as:

• time
• summary
• precipitation
• temperature
• dewpoint
• humidity
• wind
• UV index
• visibility
• ozone
• and more

but the endpoint also provides forecast:

• minutely
• hourly
• and daily

Let’s compare this to what Google thinks:

Google vs Dark Sky Current Weather Comparison

	Google	Dark Sky
Temperature	59°C	59.07°C
Humidity	61%	59%
Wind	10 MPH	12.17 MPH (converted from 5.44 Meters per second)

As you can see in the table above, the Dark Sky returns similar results to what we see from Google’s real-time weather estimates.

OpenWeatherMap

Now let’s see how OMW compares.

To begin, we’ll need to use the Current Weather Data endpoint.

Connect to the OpenWeatherMap APIOMW provides a few options when retrieving real-time weather. You can either call the API with the city string parameter or by lat-long coordinates.

We’ll try the API with the same lat/long coordinates we used with the Dark Sky API:

Not only does the OpenWeatherMap API provide the current temperature, but also what it “feels like.”

Along with temp, it also returns:

• visibility
• wind stats
• cloud data
• sunrise/sunset times
• and more

How does it compare to Dark Sky and Google?

	Google	Dark Sky	OpenWeatherMap (OMW)
Temperature	59°C	59.07°C	62.02°C
Humidity	61%	59%	62%
Wind	10 MPH	12.17 MPH (converted from 5.44 Meters per second)	11.41 MPH

As you can see here, OMW is comparable to Dark Sky and provides similar data.

When it comes to real-time/current weather data, it really depends on your preference. There are pros and cons to each one, but overall, I think OpenWeatherMap slightly takes the cake. Dark sky does provide a lot of data for any given latitude and longitude. But OpenWeatherMap provides slightly more data points such as the “feels like” temperature, cloud data, and more.

Which Service is better for Weather Forecasts?

Now let’s move onto weather forecasts.

We’ll start with Dark Sky again.

Dark Sky

To begin, we’ll be using the same API endpoint as before (GET Forecast).

This time, let’s look at the forecast data:

The Dark Sky Forecast endpoint provides a forecast by:

• minutes
• hours
• days

Let’s take a peek at each.

By the Minute

If you’re looking for really granular forecast data, Dark Sky provides a minute-by-minute forecast up to 60 minutes from the time you call this endpoint.

At a glance, you can see the forecast returns a summary of “Partly cloudy for the hour.”

For each minute, the response returns the:

• time
• precipitation intensity (Millimeters per hour)
• precipitation probability (between 0 and 1)

Looks like it’ll be a dry day.

Connect to the Dark Sky API

By the Hour

The hourly weather forecast is provided up to 48 hours from the current time.

As with the minutely response, the hourly forecast also provides a summary: “Mostly cloudy throughout the day.”

However, the hourly data is slightly more comprehensive:

For each hour, the API provides:

• time
• summary
• precipitation data
• temperature
• dew point
• humidity
• pressure
• wind data
• cloud cover
• uv index
• visibility
• ozone (Dobson units)

By the Day

The daily forecast provided is up to 7 days from the current time.

The daily weather data is a little more comprehensive than the minutely and hourly data:

For each day, you can see:

• time
• summary
• sunrise and sunset times
• moon phase
• precipitation
• temperature
• dew point
• humidity
• pressure
• wind data
• cloud cover
• uv data
• ozone
• visibility
• max and min temperatures
• and more

That’s quite a bit of data if you’re looking for minute-by-minute, hour-by-hour, and day-by-day weather forecasts.

OpenWeatherMap

On to OpenWeatherMap.

The OMW provides a 5 day / 3 hour forecast endpoint and a 16 day / daily forecast endpoint.

Let’s take a look at each.

5 day / 3 hour forecast data

We’ll start with the 5 day / 3 hour forecast data endpoint.

This endpoint provides a 5-day weather forecast for any city/location in 3-hour increments:

For each 3-hour increment, the API returns:

• Time (dt) – in unix, UTC
• weather data
  • temperature
  • “feels like”
  • pressure
  • sea level
  • ground level
  • humidity
  • and more
• cloud data (%)
• wind
  • speed
  • direction (degrees)
• rain (if any)
• snow (if any)

Compared to Dark Sky’s hourly forecast, the data that this endpoint returns is quite comparable.

Unlike the Dark Sky hourly forecast which provides data such as ozone, visibility, dew point, the OMW forecast data gives unique data such as pressure, sea level, and ground level.

In addition, OpenWeatherMap provides some supplemental data about the current location such as timezone and city populations.

16 day / daily forecast data

Another forecast API endpoint OpenWeatherMap provides is a daily (up to) 16-day weather forecast:

You can call this API by:

• city name
• lat / long
• zip code (and country)

Some optional parameters to manipulate the data include:

• cnt – (count) or amount of days in the future to forecast (no more than 17)
• units – metric or imperial
• mode – JSON or XML format (by default it returns JSON data)
• lang – language

For each day, the API returns:

• temperature
  • during the day
  • min and max
  • during the night
  • evening temp
  • morning temp
• feels like
  • day
  • night
  • evening
  • morning
• pressure
• humidity
• weather
  • conditions
  • description
• wind
  • speed
  • direction
• cloud (%)
• rain
• snow

That is a whole lot of data to work with no matter what project you’re working on.

Connect to the OpenWeatherMap APII have to say OpenWeatherMap is the overall winner in terms of forecasting data, by a slight edge. Honestly, you can’t really go wrong with using either API for predicting future weather forecasts, but OpenWeatherMap provides accurate data for slightly longer time ranges (16-day vs 7-day). However, if you’re looking for more granular data (by the hour and minute), Dark Sky is the way to go.

Which Tool is better for Historical Weather Data?

We have looked at current weather data and forecasted data, but what about historical weather data?

Dark Sky

One of the coolest features of the Dark Sky API is its “Time Machine” endpoint.

A Time Machine Request returns the observed (in the past) or forecasted (in the future) hour-by-hour weather and daily weather conditions for a particular date.

Let’s try it out.

What was the weather like last year?

Connect to the Dark Sky APIAccording to Dark Sky’s Time Machine, this time last year was a cool 53.24°C.

Not only does the Time Machine endpoint provide the same data as seen in the weather forecasts earlier, but it also has hourly data for the specified day:

For each hour, the API also returns:

• time
• temp
• apparent temperature
• dewpoint
• humidity
• pressure
• wind speed
• wind bearing
• cloud cover
• UV index

How does OMW stack up?

OpenWeatherMap

While the OpenWeatherMap API on RapidAPI doesn’t provide Historical data, you can subscribe to their historical weather APIs directly on their site, which includes:

• Historical weather API
• Statistical weather data API

Here’s what their pricing for each API/dataset looks like:

Here’s a sample API response from their historical weather API, which provides hourly data for cities:

{
  "message": "Count: 24",
  "cod": "200",
  "city_id": 4298960,
  "calctime": 0.00297316,
  "cnt": 24,
  "list": [
    {
      "dt": 1578384000,
      "main": {
        "temp": 275.45,
        "feels_like": 271.7,
        "pressure": 1014,
        "humidity": 74,
        "temp_min": 274.26,
        "temp_max": 276.48
      },
      "wind": {
        "speed": 2.16,
        "deg": 87
      },
      "clouds": {
        "all": 90
      },
      "weather": [
        {
          "id": 501,
          "main": "Rain",
          "description": "moderate rain",
          "icon": "10n"
        }
      ],
      "rain": {
        "1h": 0.9
      }
    },
    ....

and a sample from their Statistical weather data API, which allows you to get statistical data by main weather parameters for a specific date or month of the year, or for the entire year.

{
   "cod":200,
   "city_id":5400075,
   "calctime":3.105838634,
   "result":[
      {
         "month": 1,
         "day": 1,
         "temp":{
            "record_min":274.44,
            "record_max":290.45,
            "average_min":276.79,
            "average_max":286.5,
            "median":281.26,
            "mean":281.29,
            "p25":278.57,
            "p75":283.83,
            "st_dev":3.69,
            "num":136
         },
         "pressure":{
            "min":1014,
            "max":1026,
            "median":1021,
            "mean":1020.5,
            "p25":1019,
            "p75":1023,
            "st_dev":3.25,
            "num":135
         },
         "humidity":{
            "min":23,
            "max":100,
            "median":70,
            "mean":65.88,
            "p25":47,
            "p75":82,
            "st_dev":20.63,
            "num":133
         },
         "wind":{
            "min":0,
            "max":10,
            "median":1,
            "mean":2.36,
            "p25":1,
            "p75":3,
            "st_dev":2.3,
            "num":136
         },
         "precipitation":{
            "min":0,
            "max":0.3,
            "median":0,
            "mean":0.01,
            "p25":0,
            "p75":0,
            "st_dev":0.04,
            "num":136
         },
         "clouds":{
            "min":0,
            "max":90,
            "median":1,
            "mean":20.18,
            "p25":1,
            "p75":40,
            "st_dev":31.91,
            "num":136
         }
      },
      .....
   ]
}

Similarly, OpenWeatherMap provides hourly data and just as much granular data as Dark Sky including:

• temperature
• “feels like”
• pressure
• humidity
• min and max
• wind speed & direction
• clouds
• rain or snow

When it comes to historical weather data, OpenWeatherMap is again the winner here. While Dark Sky provides the hourly data for a given day, OMW provides hourly data as well as aggregated data for days, months, and years. However, using OpenWeatherMap may break the bank. Dark Sky is a great OpenWeatherMap alternative for historical weather data if you don’t need aggregated data.

Connect to the OpenWeatherMap API

Which API has the Best Overall Value?

Now let’s dive into the real question:

“Which API is going to give me the most bang for my buck?”.

Let’s look at each API.

Dark Sky

Here’s a screenshot from Dark Sky’s pricing page:

The Dark Sky weather API offers a freemium-based plan, which allows developers to call the API up to 100 times/month without incurring fees.

However, following that initial 100 API calls/month, you’ll be charged $0.001 per additional call.

Connect to the Dark Sky APINow let’s look at OMW.

OpenWeatherMap

Here’s what the OpenWeatherMap API pricing looks like on RapidAPI:

OMW’s API pricing is similar to Dark Sky’s in that they offer a freemium plan, but offers 100 free API calls/day instead of per month.

However, OpenWeatherMap’s Basic plan is rate-limited at 10 requests/minute and also has a hard limit of 100 requests/day.

In addition, they also offer a Pro paid plan for unlimited API calls for just $10/month, which is rate-limited at 100 requests/minute.

When it comes to pure value, OpenWeatherMap API on RapidAPI wins. That’s mostly because you get more API requests (100 per day vs 100 per month) for the basic freemium plan, and also OMW offers an unlimited Pro plan while you’ll end up breaking the bank with Dark Sky’s pay per request ($0.001 after the initial 100/month) if you’re scaling API calls.

Connect to the OpenWeatherMap API

Summary: OpenWeatherMap vs Dark Sky

Honestly, you can’t go wrong with each API and they might even complement each other depending on the project you’re working on.

To come up with an overall winner, we discussed:

• Overall weather data
• Real-time weather
• Comparison to other weather services (Google)
• Forecasts (by minute, hour, and day)
• Historical data
• Overall value

So if you could only pick ONE API to use for weather data, we’d have to go with the OpenWeatherMap API.

This was a hard choice, but from each of the factors discussed above, I’m sure you’d agree with us.

If you’re only making a few requests, Dark Sky might be worth it for you.

But if you’re a power user and scaling out requests, we would highly recommend OpenWeatherMap.

Here’s a table summary of what we discussed:

	Dark Sky	OpenWeatherMap
API Endpoints	2	4
Popularity	9.5/10	9.9/10
Latency	378ms	101ms
Success Rate	99%	97%
Pricing	Freemium - 100 requests/month, then $0.001 per call	Freemium - 100 requests/day (hard limit) Pro Plan - $10/month for unlimited calls.
Historical Data	Yes, via "Time Machine"	Yes, but not available on RapidAPI
Forecast Data	Minutely, hourly and daily	5 day (3 hour increments) and 16 day (daily)
	Connect to API	Connect to API

Weather API Tutorials

Whichever API you decide to use, we’d like to help you out. Here are a few tutorials by API that might help get you started.

Dark Sky Tutorials

• How to Create a Weather App with Dark Sky (Node.js)

Connect to the Dark Sky API

OpenWeatherMap Tutorials

• RapidQL Tutorial: MySQL + OpenWeatherMap + Twilio SMS
• OpenWeatherMap API and React
• OpenWeatherMap API and Node.js
• OpenWeatherMap API and Python

Connect to the OpenWeatherMap API

Now It’s Your Turn

And now it’s time to hear from you, the developers.

When it comes to OMW vs Dark Sky, which API do you prefer?

Or is there another Weather API that trumps these two for you?

Let us know in the comments below.