gu-mix.com prototype

gumix

Prototype UI for image plus voice lip sync generation with separate speech and video prompts.

Pipeline: Irodori -> LTX or audio upload -> LTX

Generate

Speech text is for Irodori. Video prompt is for LTX motion and scene content.

Base portrait image for LTX.
Optional. If present, the pipeline becomes audio upload -> LTX.
Required only when audio is not uploaded.
Always sent to LTX as the visual content prompt.
Optional. Sent to LTX as negative prompt.
Only used when audio is not uploaded.
Result is shown as mp4.