Configure the AI model and voice style used for ceremony generation.
Language Model
Select which model in LM Studio to use for generating ceremony scripts.
Current: loading...
Voice Generation
Checking TTS model...
Describe the voice. The instruction is sent directly to the TTS engine every clip. Voice can vary slightly between generations - use Preview to find a take you like, then lock it for cloning.
Current: loading...
Sounds right? Lock this preview as your clone reference - then switch to Clone Voice for consistent playback.
Pick a built-in speaker. The Voice Instruction below shapes the emotion and pacing on top of the speaker's voice - both apply.
Loading speakers...
Generate All Previews runs in the background - you can close this page and come back.
Uses a saved WAV reference sample for every clip - consistent voice. Upload your own recording or generate one via Voice Design.
No clone sample yet
Generate Sample uses the Voice Instruction from Voice Design mode to create a reference, then saves it here.
Generation Prompts
Prompts sent to LM Studio when generating phrases and jokes from AI. Edit to customize the style, count, or tone.