#voice

  1. Dessa: Block's conversation model trained only on speech

    Block's AI research lab trained a speech-to-speech model from scratch on recorded human conversation, with no text at all.

  2. MOC: Which Open-Source TTS (Voice) Model to Use

    Map of Content for picking a local, open-weights text-to-speech model as of October 2026.

  3. Stitch by Google