Skip to main content
lovino
Start free

Image to Image AI

Lovino's image to image AI makes a new picture from one reference photo and a written prompt. Restyle a portrait, change the season or add an object, and pick the model that fits. It costs 1–7 credits by model, and the 3 free daily credits cover models that need no paid plan.

1–7 credits by model · shown before you generate

  • The same woman rebuilt as a clay figure in a handmade miniature cafe
  • A smiling young woman with dark curly hair and a mustard knit sweater at a cafe table
  • The same woman as a comic book panel with bold ink lines
  • The same woman as a white marble statue in a museum hall
  • The same woman in a knit hat and scarf at a snowy winter market with string lights
  • The same woman as a Japanese woodblock print with stylised clouds

What is Image to Image AI?

Image to image AI starts from a photo you upload instead of a blank prompt. You describe the change, pick a model, and Lovino makes a new image that follows the reference. Restyle a portrait, change the season or add one object. Credits run from 1 to 7 depending on the model.

How to use Image to Image AI

Diagram: upload a reference photo, describe the change, pick an image model, and Lovino makes a new image from it. Four real results from one photo: clay figure, comic book, marble statue and winter market.
How image to image AI works: one reference photo, many new images.
  1. 01

    Upload a reference photo

    Pick the photo the new image should follow. A sharp photo with a clear subject gives the model more to hold on to.

  2. 02

    Describe what to change

    Write the change and what must stay, for example: restyle this photo as a clay figure, keep the same person, pose and composition.

  3. 03

    Pick a model

    Choose a model that takes a reference photo, such as FLUX Kontext Fast, Nano Banana 2 Lite, Seedream 4 or GPT Image 2. The credit price shows before you generate.

  4. 04

    Generate and compare

    Run it, then change the prompt or the model and run again on the same photo. Download the one you like.

One reference photo, many new images

The reference portrait and the dog, park and beach photos are AI-made test images, not real people or places. Each look is a separate run on the same portrait. The three before and after pairs are instruction edits made with FLUX Kontext Fast. Prompts are shown in full.

  • The same woman rebuilt as a clay figure in a handmade miniature cafe
    Nano Banana 2 Lite

    2 credits

    New image from the reference photo, Clay look.

    Restyle this photo: a handmade claymation character in a miniature café set, plasticine textures with tiny fingerprints, rounded shapes, soft studio light. Keep the same person, face, pose and composition. Do not add any text, letters, logos or signage.

  • A smiling young woman with dark curly hair and a mustard knit sweater at a cafe table

    The reference photo: an AI-made test portrait.

  • The same woman as a comic book panel with bold ink lines
    Nano Banana 2 Lite

    2 credits

    New image from the reference photo, Comic look.

    Restyle this photo: a bold comic book panel, thick ink outlines, halftone dots, flat vivid colours, dramatic shading, no speech bubbles. Keep the same person, face, pose and composition. Do not add any text, letters, logos or signage.

  • The same woman as a white marble statue in a museum hall
    Nano Banana 2 Lite

    2 credits

    New image from the reference photo, Marble look.

    Restyle this photo: a polished white marble statue carved from a single block, museum spotlight, subtle veins in the stone, soft shadows. Keep the same person, face, pose and composition. Do not add any text, letters, logos or signage.

  • The same woman in a knit hat and scarf at a snowy winter market with string lights
    Nano Banana 2 Lite

    2 credits

    New image from the reference photo, Winter look.

    Restyle this photo: the same scene on a snowy winter evening, gentle falling snowflakes, warm fairy lights, cosy scarf and hat, magical glow. Keep the same person, face, pose and composition. Do not add any text, letters, logos or signage.

  • The same woman as a Japanese woodblock print with stylised clouds
    Nano Banana 2 Lite

    2 credits

    New image from the reference photo, Woodblock look.

    Restyle this photo: a traditional Japanese ukiyo-e woodblock print, flat layered colours, fine carved lines, washi paper texture, stylised clouds. Keep the same person, face, pose and composition. Do not add any text, letters, logos or signage.

  • A golden-brown dog sitting on grass, looking at the camera

    Before: the dog photo. An AI-made test photo.

  • The same dog sitting on grass wearing a red bandana
    FLUX Kontext Fast

    After: add a red bandana, everything else the same.

    Put a small red bandana around the dog's neck. Keep everything else the same.

  • A green park path lined with leafy trees in summer, with a bench

    Before: summer.

  • The same park path covered in snow with frosted trees
    FLUX Kontext Fast

    After: the same scene in winter.

    Make it a snowy winter day with snow on the ground and trees. Keep the path, bench and framing the same.

  • A red beach umbrella on a sandy beach with a few people in the distance

    Before: people on the beach.

  • The same beach and red umbrella with the people removed
    FLUX Kontext Fast

    After: people removed.

    Remove the people in the background. Keep the umbrella, sand and sea exactly the same.

Models you can use here

AI image from a reference photo: restyle it

Image to image keeps what your photo shows and changes how it looks. The portrait above became a clay figure, a comic panel, a marble statue and a winter market scene. Each one is a separate run on the same photo.

The prompt does two jobs. It names the new look, and it says what to keep. The examples all end with the same line: keep the same person, face, pose and composition.

Img2img online: change one thing, keep the rest

Instruction-editing models such as FLUX Kontext Fast work differently. You describe one change, like adding a bandana, turning summer to snow or removing people, and the model leaves the rest of the photo alone. The three before and after pairs above are that kind of edit.

Pick a restyle when you want a new look for the whole picture. Pick an instruction edit when most of the photo is already right.

Which model to pick

Not every model takes a reference photo. Text-only models such as Z-Image Turbo and Imagen 4 Fast do not. The table lists models that do. Start with a low-priced one to check the idea, then run the same photo and prompt on a stronger model if you want more.

What it costs

Each run costs 1–7 credits depending on the model. The 3 free daily credits cover models that need no paid plan, which is 1 run a day at 2 credits. Images made on free credits carry a watermark. More runs need credits from a plan (from $7.99 a month). Anyone can buy a credit pack, with or without a plan. A starter pack is $2.99 for 15 credits.

Models that take a reference photo

A selection, with the lowest to highest credit price per image.

ModelCreditsNeeds a paid plan
FLUX.2 Klein1No
Nano Banana 2 Lite2No
FLUX Kontext FastShown before you generateNo
Seedream 43No
GPT Image 23–7Low quality: no. Medium and high: yes
Seedream 4.54–6Yes

What Lovino does and does not do here

Lovino doesLovino does not
ReferenceFollows your photo when you say what to keepGuarantee the same face or fine detail on every run
ModelsLets you run the same photo on several modelsTake a reference photo on text-only models
EditsChanges one thing on instruction-editing modelsPromise that everything outside your change stays identical
CostShows the credit price before you generateCharge a flat price: it depends on the model

What it can't do

  • A new image is a redraw, not a pixel edit. Faces, hands and small text can change, so compare a few runs.
  • Text-only models cannot take a reference photo. Pick one from the table.
  • Some models need a paid plan, and images made on free credits carry a watermark.
  • The model decides how closely to follow the photo. Saying what to keep in the prompt helps, but it is not a guarantee.
  • Only upload photos you have the right to use, and people who have agreed. Photos and prompts are checked, and content that breaks the rules is blocked.

Pricing

Start free. Scale anytime.

One credit system across Studio, Flow and every tool. Cancel anytime.

Save 20% on yearly plans
PlanPer monthCredits
Free$03 credits per day
Plus$7.9950 credits per month
Pro$19.99200 credits per month
Max$44.99500 credits per month
See pricing

Questions, answered.

Is image to image AI free?

Partly. Models that need no paid plan run on the 3 free daily credits. Nano Banana 2 Lite costs 2 credits, so that is 1 run a day, and images made on free credits carry a watermark. Models marked paid plan, such as Seedream 4.5, need one. Credits come from a paid plan or a credit pack.

What is image to image AI?

It makes a new image from an existing one. You upload a reference photo, describe the change, and the model draws a new picture that follows the photo. It is also called img2img.

Which model should I use for image to image?

For a quick restyle, try Nano Banana 2 Lite. For one precise change, such as adding an object or removing people, try FLUX Kontext Fast. For readable text in the new image, try GPT Image 2. Run the same photo on two models and keep the better result.

How do I keep the same face?

Say so in the prompt. The examples above all end with: keep the same person, face, pose and composition. Likeness is not guaranteed, so compare a few runs and keep the closest.

Can I use more than one reference image?

Some models can. Nano Banana 2 Lite accepts several reference images, and Lovino's Style Transfer tool sends it a photo plus a style image. The app shows what each model accepts.

How is image to image different from style transfer?

Image to image takes one photo and a prompt, and works with many models. Style Transfer takes a photo and a style image, runs on Nano Banana 2 Lite, costs 3 credits and uses paid credits only.

Can I change just one part of a photo?

Yes, with an instruction edit. Describe the one change and FLUX Kontext Fast leaves the rest. For a more targeted edit, the AI Image Editor costs 3 credits and uses paid credits only.

Can I use the new images commercially?

Images you generate are yours to use. Make sure you had the right to use the reference photo, and only use people who have agreed.

More AI tools

Related guides

Try a template

All templates

Ready to try Image to Image AI?

1–7 credits by model · shown before you generate. No purchase is made until you confirm a generation.

Try Image to Image AISee all tools