Cartesia

Cartesia makes fast, natural text-to-speech models (Sonic) and speech-to-text models, with Python and JavaScript SDKs. You send text and a voice ID and get audio back, with optional emotion, speed and volume guidance.

Website Docs Changelog Compare with others

What you can build with Cartesia

CartesiaRecipePaste

Turn text into an emotional voice clip saved as a WAV file

Generate speech from text with Cartesia Sonic 3.6, guide the emotion of the delivery, and save it as a WAV file in a few lines of Python.

Voice

Recent updatesAll updates

Nothing new in the last 30 days. For older changes, see the official changelog.