Requesting the Hume Octave model

#90
by jerhadf - opened

Hume recently released their new Octave model for expressive text to speech -- hume.ai

See also their documentation here: https://dev.hume.ai/docs/text-to-speech-tts/overview

Their model is LLM-based so it takes in voice descriptions too - and does best when it has those - but you could compare it to other models by just sending in input text without specifying a voice. See arena.hume.ai for how they're comparing to elevenlabs on just voice design.

Hi,
Thanks for reaching out! Would Hume be interested in sponsoring an API key for the Arena?

Sign up or log in to comment