The API is organized in three layers, listing available sources, datasets and series.

Querying the API can be done on a browser (requires login), setting the URL parameter token to a valid token (i.e., or including the token in the request header. Visit the Profile page (Managing API keys, login required) to obtain the token.

For example, in a CURL request:

curl "<your token>"

and equivalently:

curl -H "Authorization: Token <your token>"

Data layers

The API is structured in three layers describing the sources of data, the datasets that belong to the source, and finally the tickers in each dataset.


At the top of the API there is the sources of data, which serve as access point to the statistics of each agency. Requests to this path list all the sources available. Additionally, you can specify a source by naming its prefix or its name.

prefixSource prefixVoidYes
sourceSource nameVoidYes


For each source data are partitioned in several datasets. Datasets contain tickers that share particular features like topic, frequency or survey. Calling /api/datasets/ lists all datasets. You may append the dataset name to the path to see only one dataset (example: /api/datasets/WEO/), or filter datasets by source (example: /api/datasets/?source=IMF).

sourceSource (by prefix or name)VoidYes
datasetDataset nameVoidYes


The series path serves all the statistics. The full body of data can be filtered using several parameters.

You may request all the tickers in a dataset (example: /api/series/?dataset=pwt), and further filter the tickers by geography (example: /api/series/?dataset=weo&geography=FR).

tickerList of tickersVoidYes
datasetSeries datasetVoidYes
format(api, json, xml, csv)apiYes
tokenAuthentication token.VoidYes