Speech to Text

POST /api/v1/speech/stt

Transcribe an audio file in a supported African language.

Request

Multipart form data:

Field	Type	Required	Description
audio	file	Yes	Audio file (WAV, MP3, OGG, WebM). Max 10MB.
language	string	Yes	Language code of the audio, e.g. lug

Supported languages: lug, ach, nyn, teo, lgg, sw, en, rw, xog, myx

Response

Field	Type	Description
transcript	string	Transcribed text
detected_lang	string	Detected language code
language	string	Requested language
model_used	string	Model that processed the audio
provider	string	Provider used
latency_ms	int	Processing time in ms

Example

with open("audio.wav", "rb") as f:
    stt = client.transcribe(f.read(), language="lug")

print(stt.transcript)
print(stt.latency_ms)