PindoDocs
Web AppWeb

Speech-to-Text

Convert spoken language in an audio file into written text. STT supports Kinyarwanda, Kiswahili, and Luganda, selected with an ISO 639-1 language code in the URL.

POST/ai/stt/{lang}

Headers: Content-Type: multipart/form-data

Request

FieldTypeDescription
audiorequiredfileThe audio file to transcribe. Sent as multipart form data.
langpathLanguage code: rw, sw, or lg. Defaults to rw.

Supported audio formats: wav, wave, mp3, ogg, flac, aac, wma, webm, mp4, m4a.

Try it

Interactive PlaygroundUpload a file or record your voice to get a transcript
Access mode
Language
Audio input
Click or drag an audio file here

Supported: wav, mp3, ogg, flac, aac, wma, webm, mp4, m4a

cURL
curl -X POST "https://api.pindo.io/ai/stt/rw/public" \
     -F "audio=@/path/to/your/file.mp3"

Example request

# Public access (rate-limited, no token)
curl -X POST "https://api.pindo.io/ai/stt/rw/public" \
     -F "audio=@/path/to/your/file.mp3"

# Authenticated access
curl -X POST "https://api.pindo.io/ai/stt/rw" \
     -H "Authorization: Bearer YOUR_ACCESS_TOKEN" \
     -F "audio=@/path/to/your/file.mp3"

Response

Returns the transcribed text and a URL to the uploaded audio.

{
  "text": "...",
  "uploaded_audio_url": "path/file_name.mp3"
}