Pronunciation scoring API.
Audio + reference text → scores
Send speech audio and the sentence it should match. Prosody returns pronunciation scores, word timing, phoneme detail, and script adherence.
What it is
Prosody is an API for products that already know what the speaker was supposed to say.
Input
Audio plus reference_text.
Output
Word scores, phoneme detail, timing, and script adherence.
Access
Use the playground without a key. Request a key for API use.
Start here
- Try the playground with sample audio.
-
Read the API docs for
POST /v1/scores. - Ask for an API key.