Multiple Languages

We often get questions about multilingual support across our common APIs. This page is to help add some clarity.

Free APIs (Skills and Title)

Lightcast offers free access to our Skills and Titles APIs.

Skills API supports multiple languages including English, Spanish, and French. To see the latest list check the Get version metadata endpoint.

Titles only supports English currently.

Classification API

We are constantly working to expand the multilingual capabilities of our core Classification API. Here is a quick breakdown of where we stand. If an endpoint is not listed below you can assume it is currently English only.

Postings/Classify

The /classifications/{release}/postings/classify endpoint does not currently support non-English languages.

Skills/Extract

The Skills Extract endpoint in our Classification API supports multiple languages which can be seen by hitting the Meta endpoint /classifications/{release}/skills/extract. This endpoint should always contain the most up-to-date list.

This endpoint can have different values set for input and output, but there are some differences between what can be used for input versus output. Check the Meta endpoint for that list.

Here is a high level list of some of our currently supported languages, but please check the Meta endpoint for latest.

Inputs	Outputs
English	English (US, UK, Canada)
Czech	Czech
Danish	Danish
French (France and Canada)	Dutch
German	French
Spanish (multiple countries)	German
	Italian
	Polish
	Portuguese
	Romanian
	Spanish
	Swedish

Locales

Some endpoints take one or more locale parameters. A locale refers to the language in which a piece of text is composed, sometimes qualified by country. The Internet Engineering Task Force defines a format which they call language tags. Where they are supported, this service recognizes language tags composed of up to two subtags. The first subtag is required and consists of a lowercased two-letter code referring to the language. A second uppercased two-letter subtag indicating the associated country may be used for greater specificity. When used, the country subtag is appended to the first, separated by a hyphen.

The endpoints that take locale parameters expect them as strings in the same format as IETF language tags described above. More references:
https://datatracker.ietf.org/doc/html/rfc5646
https://datahub.io/core/language-codes
https://www.w3.org/International/articles/language-tags/