Pricing

Use https:/query.jua.ai/v1/forecasts/cost with your payload to check how much credits your request will consume.

Non-aggregated forecasts

The factors influencing the amount of credits are:

• The number of point locations (0.1 credits per location)

• Which model is used

• The forecast horizon

• The number & type of variables requested

• Number of ICs requested

Additionally, a factor of 1.2 is applied to all queries done via the API.

The total number of credits is then computed as

$$\begin{align*} \text{Credits} &= \text{API Multiplier} \times (\text{\#Locations} \cdot 0.1) \times \text{Model Multiplier} \\ &\quad \times \text{Horizon Multiplier} \times \text{Variable Multiplier} \times {\# \text{ICs}} \end{align*}$$

The Variable Multiplier increases in steps of 0.1 (or 0.2 for advanced variables such as solar or wind at 100m). For example, requesting air_temperature_at_height_level_2m results in a variable multiplier of 1.0. Requesting air_temperature_at_height_level_2m, wind_speed_at_height_level_10m and dew_point_temperature_at_height_level_2m results in a total Variable Multiplier of 1.3 .

If instead of wind_speed_at_height_level_10m you were to request wind_speed_at_height_level_100m in the last example, the total Variable Mutliplier would be 1.4.

Full Example

Assume you

• are interested in accessing 10 points

• using EPT2 (=x1.1)

• 15d forecast (=x1.3)

• are interested in air_temperature_at_height_level_2m and wind_speed_at_height_level_10m (=x1.1)

• only the latest IC

Then the total credits are

$$\text{Credits} = 1.2 \times (10 \cdot 0.1) \times 1.1 \times 1.3 \times 1.1 \times 1 = 1.89 \text{ Credits}$$

Aggregated forecasts

Aggregating forecasts on request can help save a huge amount of credits. Say you were interested in the hourly minimum and maximum wind speed in Germany for the next 7 days. Requesting the raw data (i.e. each and every point in Germany) would result in consuming 725.78 Credits. Making an aggregated request returning the hourly min & max consumes only 5.56 credits!

How are the Credits calculated for aggregated forecasts?

The total credits for aggregated forecasts are computed as the sum of

1. The data accessed for the request (i.e. the cost of computing the data)

2. The number of rows returned (i.e. transfer costs)

The first part is computed similar to the non-aggregated forecast before, but using a base-cost per point of 0.0003 instead of 0.1. The second part is simply the number of rows returned times 0.02 .

Historical data

Accessing forecast runs that are older that 90 days (i.e. an having an init_time that is more than 90 days in the past) are classified as hindcasts.

Accessing historical data is significantly cheaper than accessing the latest forecast data to enable cost-efficient back-testing backtesting using our data. We charge only 1/1000 the credits compared to the latest forecast data. More details about available hindcasts can be found here.

Use the following endpoints to get detailed information about the available forecasts & variables:

• /v1/forecasts/meta provides details about the variables per model

• /v1/forecsts/available_init_times

See Jua Query Engine API for more in-depth details about the endpoints.