1. Documentation
  2. Features
  3. Language

Language

PRE-RECORDED
STREAMING

Deepgram’s Language feature allows you to supply a BCP-47 language tag that hints at the primary spoken language of submitted audio. Deepgram defaults to US English (en-US).

Use Cases

Some examples of use cases for Language include:

  • Customers with employees who live in regions with specific dialects or accents.
  • Customers who need to ensure a specific language in their dataset is recognized.

Enable Feature

To enable Language, when you call Deepgram’s API, add a language parameter in the query string and set it to the language you would like to recognize:

language=OPTION

To transcribe audio from a file on your computer, run the following cURL command in a terminal or your favorite API client.

Be sure to replace the placeholder OPTION with your chosen language and YOUR_DEEPGRAM_API_KEY with your Deepgram API Key. You can create an API Key in the Deepgram Console.

curl
curl \
  --request POST \
  --header 'Authorization: Token YOUR_DEEPGRAM_API_KEY' \
  --header 'Content-Type: audio/wav' \
  --data-binary @youraudio.wav \
  --url 'https://api.deepgram.com/v1/listen?language=OPTION'

Language Options

Language support is optimized for the following language/model combinations:

LanguageCodeModel(s)Tier(s)
Chinesezh betageneralbase
     Chinazh-CN betageneralbase
     Taiwanzh-TW betageneralbase
Dutchnl betageneralenhanced, base
Englishengeneral
meeting
phonecall
voicemail
finance
conversationalai
video
enhanced, base
enhanced beta, base
enhanced beta, base
base
enhanced beta,base
base
base
     Australiaen-AUgeneralbase
     United Kingdomen-GBgeneralbase
     Indiaen-INgeneralbase
     New Zealanden-NZgeneralbase
     United Statesen-USgeneral
meeting
phonecall
voicemail
finance
conversationalai
video
enhanced, base
enhanced beta, base
enhanced beta, base
base
enhanced beta, base
base
base
Flemishnl betageneralenhanced, base
Frenchfrgeneralbase
     Canadafr-CAgeneralbase
Germandegeneralbase
Hindihigeneralbase
     Roman Scripthi-Latn betageneralbase
Indonesianid betageneralbase
Italianit betageneralenhanced, base
Japaneseja betageneralenhanced, base
Koreanko betageneralenhanced, base
Polishpl betageneralbase
Portugueseptgeneralenhanced beta, base
     Brazilpt-BRgeneralenhanced beta, base
     Portugalpt-PTgeneralenhanced beta, base
Russianrugeneralbase
Spanishesgeneralenhanced beta, base
     Latin Americaes-419generalbase
Swedishsv betageneralbase
Turkishtrgeneralbase
Ukrainianuk betageneralbase

Results

Once applied, results will appear in the transcript.

language=fr

Transcript
sur votre personnage un drogué ça passe par plein de phases la défense la descente le manque la honte la colère la dépression quand on se rend les tics du visage la mâchoire les larmes le c'était quoi le plus difficile à jouer si tu le fais que si quelqu'un devrait et pas spécifique non plus un détail mais le fait que c'est c'est un homme vivant et même même indépendant des drogues ou la c'est leur vie c'est leur famille

By default, Deepgram applies its Base tier general AI model, which is a good, general-purpose model for everyday situations. To learn more about the customization possible with Deepgram's API, check out the Deepgram API Reference.

FEEDBACK