Generation modes

    Image-to-image

    Also called Img2img.

    Image-to-image takes a picture as its primary input and returns a changed picture — restyled, corrected or varied — instead of inventing one from nothing.

    The input does two things at once: it supplies content and it supplies structure. How much of that structure survives is the whole game, and it is governed by a strength setting — low values nudge the picture, high values keep only the rough layout and repaint the rest.

    This is the mode behind most "make it look like" work: turn a phone photo into a studio shot, restyle a frame into illustration, produce six variants of one hero image that still read as the same composition. It is also the mode behind fixes, because a small strength value is a targeted correction rather than a fresh roll of the dice.

    Masked variants of the same idea get their own names — inpainting when you protect everything except a region, outpainting when you push past the original edges.

    In practice

    • Strength is the dial that matters; prompt wording barely registers if strength is near zero.
    • Structure survives better than texture — layout and pose persist long after surface detail is gone.
    • Running the output back in as the next input compounds drift quickly.

    Image-to-image models

    Catalog entries that take a picture in and hand a changed picture back. 30 of the 296 models in the Versely catalog qualify.

    Browse all 10 spec pages for full settings, resolutions and credit costs.

    The mistake to avoid

    Expecting an identity to survive a high strength value. Past roughly the halfway mark the face is being redrawn, not edited, and it will no longer be the same person.

    Where you will run into it

    Related terms

    The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.