Speech-to-Text

Supported Languages

List all supported STT languages with accuracy and feature availability.

GET/api/stt/languages

Returns all supported languages with their codes, scripts, and KenLM availability.

No parameters for this endpoint.


Response

[
  {
    "code": "hi",
    "name": "Hindi",
    "script": "Devanagari",
    "region": "indian-scheduled",
    "kenlm": true
  },
  {
    "code": "en",
    "name": "English",
    "script": "Latin",
    "region": "english",
    "kenlm": true
  }
]

Indian Scheduled Languages (22)

All 22 constitutionally recognized languages of India with KenLM accuracy boost:

CodeLanguageScriptKenLM
hiHindiDevanagariYes
bnBengaliBengaliYes
teTeluguTeluguYes
mrMarathiDevanagariYes
taTamilTamilYes
urUrduArabicYes
guGujaratiGujaratiYes
knKannadaKannadaYes
mlMalayalamMalayalamYes
orOdiaOdiaYes
paPunjabiGurmukhiYes
asAssameseBengaliYes
maiMaithiliDevanagariYes
saSanskritDevanagariYes
neNepaliDevanagariYes
sdSindhiArabicYes
kokKonkaniDevanagariYes
doiDogriDevanagariYes
mniManipuriBengaliYes
brxBodoDevanagariYes
satSantaliOl Chiki
ksKashmiriArabic

International Languages

Vega supports 200+ languages via the CTC model. Dhara supports 200+ via Whisper.

Popular international languages: English (en), Spanish (es), French (fr), German (de), Chinese (zh), Japanese (ja), Korean (ko), Arabic (ar), Russian (ru), Portuguese (pt).

On this page