About 30 seconds an image
Upload, describe the change and press Remake Image. Most image to image runs finish in about half a minute, so you can try several ideas on the same picture in a few minutes.
Your picture, your words
Image to Image AI: Remake Your Picture from One Sentence
Image2Layer's image to image AI turns your picture and one sentence into a new version of the whole image, free to try. Your picture sets the layout, your words the change, and a PNG is ready in about 30 seconds.
- Free to try
- One sentence, no masks
- 6 one-click ideas
- PNG in about 30 seconds
1. Upload an image
Sexual or explicit content isn't allowed. Images and results that contain it are blocked automatically.
No image handy? See a sample
2. Describe the new image
The layout follows your picture.
Ideas
3. Remake Image
10 credits per image · Failed runs never use credits
Real result of this tool
Upload your own image to get a PNG.
One picture, four sentences
What Image to Image AI Does
Image to image makes a new version of a picture you upload, following a short description you write. Your picture sets the layout, the camera angle and where everything stands, and your words decide what the new image looks like. It works like an AI photo editor with a prompt instead of tools: change the time of day, the season, the weather or the light, turn a sketch into a photo, or give a room a new style, with no brush, mask or layers.
The tabs show one street remade four times, each from a single sentence on Closely. Because image to image remakes the whole picture, it is the right tool when the whole scene should change. For preset art styles with no description, use photo to painting; to change only a brushed area, use generative fill; to change the words on a sign or a label, use edit text in image. Afterwards, press Split into layers to open the new image in image to layers.
- Day to night and golden hour
- Snow, rain and new seasons
- Sketch or model to photo
- New interior styles
- Clay, film and cinematic looks
- Closely, Balanced or Loosely
New imageWhy Use This Image to Image Generator
Six reasons to edit images with AI here instead of rebuilding a picture by hand.
Your composition stays
On Closely the layout, camera angle, positions and proportions are kept and only what you describe changes, so the new image lines up with the one you uploaded.
Six ideas, one click each
Night scene, Snowy winter, Golden hour, Rainy day, Sketch to photo and New interior style fill the box with a ready sentence that you can edit before you run it.
Choose how closely it follows
Closely, Balanced or Loosely decides how much of your picture carries over, from a careful change of light to a fresh image that only borrows the layout.
Six looks to finish it
Keep style, or render the result as a Photo, a 3D render, a Cinematic film still, a Clay figure scene or Vintage film.
10 credits, failed runs free
Each image costs 10 credits. Credit packs need no subscription, and a run that fails never uses credits.
How to Use Image to Image in 3 Steps
How to use image to image with no editing skills: three steps from your picture to a new PNG, with no brush and no layers.
JPGPNGWebPup to 24 MB
Image to Image That Keeps Your Composition
Three things this image to image AI editor does on example pictures. Drag the divider, move the lens or watch each example.
CloselyLooselyClosely, Balanced or Loosely: Decide How Much Changes
Follow the original sets how much of your picture the new image keeps. Closely keeps the layout, camera angle and positions and changes only what you describe. Balanced keeps the subject and composition recognizable while the style and details move, and Loosely treats your picture as loose inspiration. This lighthouse got the same sentence twice, "a stormy night with crashing waves and a bright lighthouse beam": drag the divider to compare Closely and Loosely.
Choose how closely
Move to see the modelSketch to Photo AI for Drawings, Models and Renders
Upload a pencil sketch, a 3D render or a foam model, describe the real thing and pick Photo under Style, and the AI renders it as a photograph in the same layout. This white architecture model became a timber house on a sunny street: move the lens to see the model underneath. Where a sketch leaves things out, the AI fills them in, so check doors, windows and small parts.
Turn a sketch into a photo
AfterClay, Film and Cinematic Looks in One Run
Style decides how the whole result is rendered. Keep style leaves the look of your picture alone, while Photo, 3D render, Cinematic, Clay figure and Vintage film each give it one finished look. This seaside boardwalk got Vintage film: warm faded colors and soft grain, with every bench and lamp post where it was. The zoom shows the grain up close. For a larger file, enlarge the result with the image upscaler.
Pick a lookImage to Image Before and After
Six example pictures made for this page, each remade by this image to image AI tool with the description and settings in its note. Hover a card to wipe between them, or open it to compare closely.
Toy boat to real sailboat
A wooden toy sailboat in a park puddle became a real sailboat on a calm sea at sunrise, with the same hull colors and sails.
Backyard to garden
A bare backyard with patchy grass got a stone path and flower beds, seen from the same corner of the house.
3D render to product photo
A plain gray render of a desk lamp became a product photo on an oak desk in soft daylight.
Corner cafe as a 3D render
A corner cafe on a sunny morning turned into a polished 3D render with soft light, every table and awning in place.
Desert road at blue hour
A desert highway at noon became a cinematic blue hour scene with fog over the hills, the road in the same place.
Summer to autumn, with a person
A woman reading on a park bench moved from summer to autumn with falling leaves, her pose and the bench kept on Closely.
Image to Image Prompt Tips That Work
Four habits for a description the AI follows, with image to image prompt examples you can copy.
Describe the new image, not the old one
The AI already sees your picture, so spend your words on the result: "the same street at night with warm lights", "this room in a modern Scandinavian style". Any AI image editor with a prompt works best this way. Start with "the same scene" when the layout should stay.
One big change per run
Night, snow and a new style in one sentence pull in different directions. Edit images with prompts one step at a time: make one change, press Edit the result, then describe the next change.
Match Follow the original to the job
Closely suits rooms, buildings, products and pictures with people, where things must stay in place. Balanced suits a new look, and Loosely a fresh idea that only borrows the layout. Faces can change, so check them; for a new haircut on the same face, use the AI hairstyle changer.
Start from a clear, well-lit picture
The new image follows the shapes it can see, so a sharp photo or a sketch with clean lines gives the best result. Uploads are scaled to fit 2048 pixels on the longest side, and the result keeps the shape of your upload.
Image to Image vs Photoshop and a New Photo Shoot
This image to image AI generator next to editing by hand in Photoshop and booking a new shoot for the shot you need.
- Time per imageImage2Layer:About 30 secondsPhotoshop by hand:Often hours for a day to night editA new photo shoot:Days, plus waiting for the light or the season
- CostImage2Layer:10 credits an imagePhotoshop by hand:A Photoshop subscriptionA new photo shoot:A photographer, travel and props
- Skill neededImage2Layer:None: write one sentencePhotoshop by hand:Masks, layers, lighting and practiceA new photo shoot:A photographer and the right weather
- What can changeImage2Layer:Time, season, weather, light, style and materialsPhotoshop by hand:Anything, painted by handA new photo shoot:Whatever the day gives you
- Same compositionImage2Layer:Kept on Closely, check small detailsPhotoshop by hand:Exact, you edit the originalA new photo shoot:Hard to repeat the same angle
- What you getImage2Layer:A PNG of about 1,200 pxPhotoshop by hand:A file at any size you chooseA new photo shoot:Full size camera photos
Photoshop is a trademark of Adobe. Image2Layer is not affiliated with Adobe.
What People Remake with Image to Image
Ways to change a photo with an AI prompt: homes, designs, products, places and stories.
Interiors and real estate
Show a room in a new style, or use it as a day to night photo AI for a listing at dusk. Label a restyled room as a concept, not the room as it is today.
Architects and designers
Turn a sketch, a foam model or a plain 3D render into a realistic photo for a mood board or an early client review.
Product ideas
Turn a product render into a photo-style mockup, or let the image to image generator show the same product in new light.
Travel and places
Show a town in snow, a beach at golden hour or a street on a rainy night for a blog post, a slideshow or a poster.
Storyboards and concept art
Edit images with an AI prompt to give one location several moods for a film, a game or a story: day, night, fog and rain, with the same layout each time.
Design projects
Split the new image into editable layers with image to PSD, or make a matching poster from a short description with the AI poster generator.
Image to Image FAQ
Answers about descriptions, settings, faces, price, size and privacy.
It makes a new version of a picture you upload, following a sentence you write. Text to image starts from words alone, while image to image AI with a text prompt starts from your picture, so the layout and the subject carry over and the time, season, light or look changes.
Upload a picture, type the new image into the description box, choose Closely, Balanced or Loosely and press Remake Image. About 30 seconds later the result is ready to compare with the slider and download as a PNG. The description is required: an empty box asks you to describe the image you want.
Yes, once. Each account gets one free run across Image2Layer's AI tools: sign in, press Try it free, and you get a watermarked result with no credits needed. The watermark disappears as soon as you buy any plan or credit pack, and after that each image costs 10 credits.
A good prompt says what the new image should look like, in one sentence. The six ideas are good starting points, such as "the same scene at night with warm lights" or "redecorate this room in a modern Scandinavian style". Starting with "the same scene" helps keep the layout, and up to 600 characters fit in the box.
They set how closely the new image follows your picture. Closely keeps the composition, camera angle, positions and proportions and changes only what you describe. Balanced keeps the main subject and the layout recognizable while the style and details change. Loosely uses your picture as loose inspiration for a new one.
They give the whole result one finished look. Keep style leaves the look of your picture as it is, Photo makes a realistic photograph, 3D render a polished render with soft light, Cinematic a film still with dramatic light and color grading, Clay figure a scene of glossy plasticine, and Vintage film warm faded colors with soft grain.
No, not reliably. The whole picture is redrawn, so faces can change, most of all on Balanced and Loosely. Closely keeps people in place and often close to how they looked, but check every face before you use the result. To try a new haircut on the same face, the AI hairstyle changer is the better tool.
Yes. Upload a pencil sketch, a line drawing, a foam model or a plain 3D render, press Sketch to photo or describe the real thing, and pick Photo under Style. It works as a drawing to photo AI: lines become surfaces, light and texture, and the AI fills in what the sketch leaves out, so check the details.
Yes. Press Night scene or write something like "the same street at night with warm lights", and keep Closely so the buildings stay where they are. The same works for snow, rain, fog, autumn leaves or golden hour light.
Not exactly. On Closely only what you describe should change, but image to image redraws the whole picture, so small details elsewhere can shift. To change one area and leave the rest untouched, brush it in generative fill; to change words, use edit text in image; to take out an object, use the object remover.
No. Follow the original, with Closely, Balanced and Loosely, takes the place of a strength number, and you write what you want rather than what to avoid. Every run starts fresh, so the same sentence gives a slightly different image each time: keep the one you like best.
No. One picture goes in and one new image comes out, in the same shape as your upload. To give a result more room on the sides afterwards, use the AI image extender.
Each image costs 10 credits, and a run that fails never uses credits. Credit packs need no subscription, and results paid with credits have no watermark. The pricing page lists the current plans and packs.
It comes back as a PNG of about 1,200 pixels on the longest side, in the shape of your upload. That suits screens, posts and mood boards; for print, enlarge it with the image upscaler. Uploads up to 24 MB are accepted.
Your uploads and results are visible only to your account, and every upload and description passes our content checks. Only edit pictures you own or have permission to use, and never make misleading pictures of real people.
More Tools to Edit Images with Prompts
Image tools for one area, words, clothes, hair or the next step, when you want to modify images with AI in another way.
Keep Editing After Image to Image
Four tools for what usually comes next.
Describe it, remake it
Remake Your Picture with Image to Image
Upload a picture, write one sentence and get a new version in about 30 seconds. Image to image online starts free with an account.
Remake a picture