Speech playground

Synthesise speech from text using any speech-capable model. Calls hit /v1/audio/speech and are billed per input character.