Settings and concepts

Terms in the app

• Service: the running local engine or API connection.
• Model: the selected speech, recognition, sound or language model.
• Voice: a prebuilt voice, supplied reference or saved synthetic voice.
• Generation run: a group of speech takes made with a recorded set of inputs and settings.
• Artifact: a saved result from a processing stage.
• Work item: a background operation with its own status.

Keep choices separate

Language coverage, voice cloning, delivery instructions and timing support are separate capabilities. A model can support one without supporting another. The following pages explain how these choices fit together.