Text-to-Image vs Image-to-Image: When to Use Reference Images

The two generation modes

• Text-to-image: the model generates from scratch using only words. Maximum freedom — best when you have no source material and want fresh ideas.
• Image-to-image / reference editing: the model builds on one or more uploaded references. Maximum consistency — best when a subject or style must be preserved.

When to use text-to-image

• Article headers and concept illustrations with no fixed assets
• Fresh creative directions unconstrained by existing images
• Rapid exploration of multiple options

When to use image-to-image

• E-commerce product shots: the real product must be reproduced
• Brand series: style and logo must stay unified
• Subject continuity: the same person or object across many images

For role-by-role reference usage, see reference images and brand consistency.

The combo is strongest

In real work the winning recipe is usually "text description + style reference": words define content, the reference defines tone. Peituka supports up to 4 combined references.

Try both

Open Peituka, generate once from text alone, then add a reference and compare the difference.