AyaSpeech Logo

Pipelines

Chain models together in a single request. Pipelines reduce latency by avoiding multiple round trips.

ASR → MT (Transcribe & Translate)

POST /transcribe_and_translate

Upload audio → get transcription in Twi → get English translation. All in one request.

curl -X POST "https://api.ayaspeech.ayadata.ai/transcribe_and_translate" \
  -H "x-api-token: your_api_token" \
  -F "audio_file=@recording.wav"

MT → TTS (Translate & Speak)

POST /translate_and_synthesize

Send English text → get Twi translation → get spoken audio. Returns a WAV file.

curl -X POST "https://api.ayaspeech.ayadata.ai/translate_and_synthesize" \
  -H "x-api-token: your_api_token" \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello world", "source_lang_name": "English", "voice": "female"}' \
  --output output.wav