HeyGen Voice STS carries a spoken take onto an avatar

HeyGen's Voice STS carries a spoken performance onto an avatar. HeyGen says it preserves pauses, emphasis, and whispers; a test page is live on its platform.

In Studio, the workflow is called Voice Mirroring. Record there or upload an MP3 or WAV up to 100 MB, then turn mirroring on. The avatar's selected voice follows your timing, and HeyGen transcribes the recording so you can check the words.

The Speech to Speech API costs $0.12 per minute and returns a WAV. For avatar lip sync, pass that audio into a video request. Professional voice clones aren't supported on the API yet.