AI Generation

Image to Image (i2i) Models, Explained

Also called img2img, image to image translation, what is image to image, reference image

Image to image, usually shortened to i2i, is the generation task where an input image conditions the output. The model rewrites an existing picture instead of starting from noise alone, so composition, color, and pose carry over to whatever degree the strength setting allows.

What changes when you add an image

A text-only generation has to invent structure. Once you supply an image, structure is already decided, and the model's job narrows to reinterpreting it. This is why i2i is the workhorse of production work while text-only generation is the ideation tool: you can predict roughly what will come back.

Under the hood, most implementations add noise to your image and then denoise it toward the prompt. How much noise gets added is the strength setting, and it is the difference between a color grade and a completely new picture.

The strength dial is most of the interface

StrengthWhat survivesUse for
Low (0.1 to 0.3)Almost everythingGrades, small cleanups, texture passes
Medium (0.4 to 0.6)Layout, pose, paletteRestyling, medium changes, look development
High (0.7 to 0.9)A rough compositionLoose reinterpretation, concept variants

Everyone's first mistake is running high strength and being surprised the source is gone. Everyone's second mistake is running low strength and concluding the model cannot restyle. Find the value where structure holds and style moves, note it, and reuse it for the whole shot.

Picking the right task

  • Change the whole frame's look. That is i2i.
  • Change one object or region. Inpainting, with a mask.
  • Apply another picture's aesthetic. Style transfer, which is i2i with a style reference rather than a text instruction.
  • Extend the frame outward. Outpainting.
  • Fix the wrong thing at the wrong resolution. Do the repair first, then upscale.

Choosing the narrowest task that can do the job is the single most useful habit in this workflow. Every extra pixel you allow the model to touch is a pixel it can get wrong.

Multiple reference images

Newer models accept several images at once, each playing a role: one for composition, one for a character's identity, one for a product or garment. This is where image to image stopped being a style tool and became an assembly tool. Two rules make it work:

  1. Say what each image is for in the prompt. Without that, the model guesses, and it usually guesses that you wanted an average.
  2. Keep the references clean and cropped tight on the thing that matters. A busy reference contributes its clutter along with its subject.

Prompting an i2i pass

Describe the change, not the picture. The picture is already in front of the model, and re-describing it competes with the pixels you supplied.

  • repaint as hand-inked comic art, flat shadows, keep pose and camera angle
  • ⚠️ a woman standing by a window in a blue dress restates what the source already fixes

Two failure modes to expect. Text and logos degrade on every pass, because they are re-synthesized each time; mask them out if they matter. And quality drifts if you chain passes, since each run inherits the previous run's artifacts. When you find yourself on the fourth pass over the same file, go back to the original and do the work in one considered step instead.

Models that support this

Pulled from the live ZOOOP model catalog, so this list stays current as new models ship.

See all 25 models

The prompt for this

A starting point that reliably produces the effect. Adjust the subject and setting; keep the technical clauses.

Repaint this photo as a hand-inked comic panel, keep the exact pose and camera angle, flat ink shadows, no color

Try Image-to-Image (i2i) yourself

Open the generator with a starting point already filled in.

Frequently asked questions

What is image to image, in one sentence?
It is generation with a picture as part of the input, so the model modifies what you gave it rather than inventing a scene from scratch. That single change is what makes results predictable enough to iterate on.
What is the difference between i2i and inpainting?
An image to image pass can rewrite every pixel in the frame, guided by the source. Inpainting restricts changes to a mask you draw and leaves everything outside it untouched. If your goal is one object or one region, inpainting is the correct task and i2i is a blunt instrument.
Why does the output ignore my input image?
The strength setting is too high. At high strength the model is allowed to depart so far from your source that only a vague color layout survives. Drop it until the structure returns, then raise it in small steps until the style changes as much as you want.
Is i2i the same as image to image translation?
Image to image translation is the older research term for mapping one visual domain onto another, such as sketch to photo or day to night. Today's models do that and much more from a text instruction, so the phrases are used interchangeably in practice.
How many reference images should I supply?
One for structure. Add a second only when it carries a different kind of information, such as a character's face or a product's exact packaging. Beyond three the model tends to average them, and you get something that resembles none of your references.
Will an image to image pass keep the same face?
Not by default. Identity is one of the first things to drift as strength rises. Either keep strength low, or mask the face and run the pass on the rest of the frame, which is a more reliable way to preserve a likeness than any prompt wording.

Related terms