Image to Image AI Generator

Add a photo or drawing and say what you want. Canvas creates a new image based on yours, or changes your image itself the way you describe.

A glass perfume bottle standing on a wet black stone, light mist behind it, a single warm light from the left, thin reflections on the glass. Keep the bottle's shape, label and color exactly as in the photo I added and change only the scene. Magazine ad mood, close-up, photorealistic, simple composition, no extra text in the image.

Blue parts are the places to fill in with your own details. Tab selects the next one, Enter sends, Shift and Enter starts a new line.

This tool runs on Entro Canvas.

Example results

Two ways to use your image

Use as reference

The new image is made by looking at yours. Every image you add gets a role: General guides composition and the overall look, Style takes only color and light, and Subject keeps a person, product or logo exactly.

Car prompts: 10 ready-to-use prompts for Gemini and ChatGPT

Edit

Your image itself is changed. Image 1 is the one that gets edited; write things like "make the background white" or "put the product on marble". Other images become an element to add (a logo, a product) or a style to apply.

Cinematic prompts: 10 ready-to-use prompts for ChatGPT and Gemini

Up to 4 images per run

Pick PNG, JPEG or WebP files with Add image, drag them onto the box or paste them. Gallery cards also have Use as reference and Edit, so you can carry on from an image you made earlier.

ChatGPT caricature prompts: 10 ready-to-use prompts for ChatGPT and Gemini

Put your product in new scenes

Add a plain photo of your product with the Subject role and describe the scene: on a beach towel, among snowflakes, on a wooden shelf. The product stays the same while the setting changes. Ask for four options and keep the one that looks most natural.

Birthday prompts: 10 ready-to-use prompts for Gemini and ChatGPT

Carry the mood of one image into another

Add a photo you like with the Style role and its colors and light are taken while its content is left behind. Use the same style image across a campaign and every visual stays in one tone.

Oil painting prompts: 10 ready-to-use prompts for ChatGPT and Gemini

When image to image helps

  • Add a product photo from your phone, choose Edit and write "make the background plain white". You get a clean image for a listing or catalog, and the edited image keeps your photo's ratio.

Flower bouquet prompt: realistic flower images and photos

Examples made from a starting image

How to create a new image from an image with AI

  1. 1

    Add your image

    In Canvas, press Add image next to the prompt box and choose up to 4 images. To start from an image in your gallery, use Use as reference or Edit on its card.

  2. 2

    Choose how it is used

    With Reference, give each image a General, Style or Subject role. With Edit, image 1 is changed; write what should change in short, plain words.

  3. 3

    Generate and compare

    Press Generate. The result lands in your gallery as a separate run and your source image stays as it was. If it is not right, change one thing in the prompt and try again.

Chibi prompt: 10 ready-to-use chibi prompts for ChatGPT and Gemini

Frequently asked questions

How many images can I add, and which files work?

You can add up to 4 images per run. PNG, JPEG and WebP are supported. Large photos are resized in your browser before upload so the longest side is 2048 pixels, which means you can add a phone photo as it is.

What is the difference between Reference and Edit?

With Reference, Canvas looks at the image you added and makes a brand new one, taking composition, style or subject according to the role you chose. With Edit, image 1 itself changes and the rest stays as close to the original as possible. Reference suits new scenes; Edit suits small changes.

Will my product look the same?

When you add a product photo with the Subject role, Canvas tries to keep the product exactly. Small label text or fine patterns can still shift, so check the result up close. If a detail is off, open Edit on the card and ask it to fix only that detail.

How many credits does an edit cost?

Each edit creates a new image and costs one image on the model you choose: 2 credits on Auto and 1 on EntroGen Flash. With a premium model the cost comes from your balance. The Edit window shows the model and the cost before you press the button.

Does every model accept images?

No. MAI Image 2.6 takes no image input and cannot edit. If you pick it and add images, the screen warns you, and edits are done with Auto instead. Auto, EntroGen Flash, Nano Banana Pro and FLUX.2 Pro all accept images.

Can I use a photo with a person in it?

Use your own photo or one of someone who gave you permission. If the prompt or image breaks the usage rules, the run does not go ahead and no credits are taken.

What ratio does an edited image get?

In Edit mode the output is set to the aspect ratio closest to the photo you are editing, so a portrait phone photo stays portrait. In Reference mode you pick the ratio yourself from ten options, from square to 21:9.

Can I add images on the Free plan?

Yes. Add image, Reference and Edit are all available on the Free plan. The daily upload limit is lower than on paid plans, and the screen tells you how much you have used if you reach it. Premium models and Magic Prompt are on paid plans; Magic Prompt is not used for edits anyway.

How do I carry on from an image in my gallery?

Each image card has two buttons: Use as reference adds the image next to the prompt box as a reference, and Edit opens the edit window. If you switch to the prompt box with More options in that window, you can add other images to the edit and choose their roles.

More AI tools

  • Car prompts: 10 ready-to-use prompts for Gemini and ChatGPT

    Text to Image Generator

    Describe a scene and Canvas turns it into 1 to 4 images. Ten aspect ratios, 16 style presets and monthly image credits on the Free plan.

  • Flower bouquet prompt: realistic flower images and photos

    AI Design Generator with Editable Designs

    Describe a post, invite or promo and Canvas designs it; then change any detail by writing. Each edit is a new version and the old one stays in your gallery.

  • Cartoon prompts: 10 ready-to-use prompts for ChatGPT and Gemini

    AI 3D Avatar Generator

    Turn your photo or a description into a 3D animated-movie style avatar image for profile pictures and team pages. You get a PNG or JPG image.

  • Movie poster prompt: cinematic AI movie poster design

    AI Poster Generator

    Describe your event, headline and style, and Canvas draws posters and flyers for you. Vintage poster, art deco and pop art presets, plus 4K downloads.

  • Chibi prompt: 10 ready-to-use chibi prompts for ChatGPT and Gemini

    AI 3D Character Creator

    Describe a mascot or character, and Canvas draws a 3D render-style design, front, side and back views and expression sets as images. Start free.

  • Restaurant Logo Prompt: 8 Ready-to-Use Prompts for ChatGPT

    AI Logo Generator

    Type your brand name and style, and Canvas draws four logo options in one go. A logo template, flat vector presets and your own color codes. Start free.

You describe it, Entrobase builds it.

Create with Entrobase