Lesson Content

Style reference in Leonardo AI.

Leonardo AI includes features for style-reference-guided generation – using a reference image to extract and apply its visual style to new content generated from the prompt. The style extraction focuses on the aesthetic character (color treatment, rendering style, texture quality) rather than the content of the reference, enabling new subject content in the reference's visual style.

Combining style reference with prompts.

Style reference and prompt work together: the style reference provides the aesthetic framework; the prompt provides the subject and content direction. For the strongest style transfer effect, use a reference that clearly represents the target style – high contrast to the current default aesthetic, with distinctive style characteristics that are clearly readable.

Evaluating style reference results.

After applying a style reference, evaluate whether the aesthetic transfer is working: does the generated output have the color character of the reference? the rendering style? the texture quality? If transfer is weak, test a different reference with more distinctive style characteristics. If transfer is overriding the prompt content too strongly, adjust the style strength.

Building a style-consistent series.

For a series: establish one verified style reference image, confirm it produces the intended aesthetic transfer on a test prompt, then apply it consistently across all series generations. The result: varied subject content in a consistent aesthetic style – the foundation of visual identity for a content series or brand.