Language and region support for the Text Analytics API

  • 2 minutes to read

This article explains which languages are supported for each operation: sentiment analysis, key phrase extraction, language detection and named entity recognition.

Language Detection

The Text Analytics API can detect a wide range of languages, variants, dialects, and some regional/cultural languages. Language Detection returns the "script" of a language. For instance, for the phrase "I have a dog" it will return en instead of en-US. The only special case is Chinese, where the language detection capability will return zh_CHS or zh_CHT if it can determine the script given the text provided. In situations where a specific script cannot be identified for a Chinese document, it will return simply zh.

We don't publish the exact list of languages for this feature, but it can detect a wide range of languages, variants, dialects, and some regional/cultural languages.

If you have content expressed in a less frequently used language, you can try Language Detection to see if it returns a code. The response for languages that cannot be detected is unknown.

Sentiment Analysis, Key Phrase Extraction, and Named Entity Recognition

For sentiment analysis, key phrase extraction, and entity recognition, the list of supported languages is more selective as the analyzers are refined to accommodate the linguistic rules of additional languages. Support for the full set of entity types is currently limited to the following languages:

  • English
  • Chinese-Simplified
  • French
  • German
  • Spanish

Only the Person, Location and Organization named entities are returned for the other languages.

Language list and status

Language support is initially rolled out in preview, graduating to generally available (GA) status, independently of each other and of the Text Analytics service overall. It's possible for languages to remain in preview, even while Text Analytics API transitions to generally available.

Language Language code Sentiment Key phrases Named Entity Recognition Notes
Arabic ar ✔ *
Czech cs ✔ *
Chinese-Simplified zh-hans ✔ *** ✔
Chinese-Traditional zh-hant ✔ ***
Danish da ✔ * ✔ ✔ *
Dutch nl ✔ * ✔ ✔ *
English en ✔ *** ✔ ✔ **
Finnish fi ✔ * ✔ ✔ *
French fr ✔ *** ✔ ✔
German de ✔ * ✔ ✔
Greek el ✔ *
Hungarian hu ✔ *
Italian it ✔ *** ✔ ✔ *
Japanese ja ✔ *** ✔ ✔ *
Korean ko ✔ ✔ *
Norwegian (Bokmål) no ✔ * ✔ ✔ *
Polish pl ✔ * ✔ ✔ *
Portuguese (Portugal) pt-PT ✔ ✔ ✔ * pt also accepted
Portuguese (Brazil) pt-BR ✔ ✔ *
Russian ru ✔ * ✔ ✔ *
Spanish es ✔ ✔ ✔ **
Swedish sv ✔ * ✔ ✔ *
Turkish tr ✔ * ✔ *

* Language support is in preview

** Named Entity Recognition and Entity linking are both available for this language.

*** Available in Sentiment Analysis V3 Public Preview

See also

Cognitive Services Documentation page
Cognitive Services Product page