Gpt-4o-mini-transcribe rejects valid ISO-639-1 code eu

API Bugs
, , ,
1

When you pass ‘eu’ as a language in stt api got this error:
“errMessage”: “400 Language code ‘eu’ is not recognized. Try adding the language name to your prompt.”
But it’s valid.

2

Welcome to the community!

Would you please clarify which language you’re referring to when you write “eu”? It will be easier to help from there.

3

Thanks! “eu” stands for Basque (Euskara). It is a valid ISO 639-1 language code. You can see it listed on Wikipedia here: wikipedia: /wiki/List_of_ISO_639_language_codes

According to the error message, the API itself isn’t recognizing the short code eu. As a workaround, you can try passing the language name “Basque” or “Euskara” directly into the initial prompt parameter of the STT API call to help the model identify the language.

4

Thank you for clarifying!

And yes, I would recommend that you test that and see if it works.