Lyvia 6

About

Lyvia 6 is a next-generation image model offering precise control over generation through denoise strength adjustment. It supports a single reference image as a starting point, face swap for consistent character identity, and outputs at up to 1280×720. A Pro-exclusive model with very low credit cost.

How to prompt

• Use denoise strength to control how much the model deviates from your reference image — lower values stay closer, higher values are more creative
• Provide a start image for guided generation: the model uses it as a composition and style anchor
• Use face swap to maintain a consistent face across a series of images
• Keep prompts descriptive but focused — one clear subject per image works best