A new color, the same shirt
“Make the linen shirt blue. Keep its cut and folds, the person, and the room unchanged.”
The garment changes color while its shape and the surrounding room stay recognizable.
Select the shirt rather than the whole person.
AI PHOTO EDITOR
Describe what you want to change—or create an image from scratch. Edit selected areas, compare versions, and keep your original.
Basic edits are free. AI images start at 10 credits with these settings. Review the price on the canvas before generating.
One clear request. A visible change.
Change a color, clear a distraction, or build a new scene. Compare the whole image before choosing your next edit.
“Make the linen shirt blue. Keep its cut and folds, the person, and the room unchanged.”
The garment changes color while its shape and the surrounding room stay recognizable.
Select the shirt rather than the whole person.
“Remove the passerby in the red coat on the right. Keep the traveler and the street unchanged.”
The main subject stays in place while the background person is removed.
Include the unwanted person’s shadow in the area to remove.
“Replace the white table lamp with a slender brass lamp and a white linen shade. Keep the room and table as they are.”
The replacement fits the same space without changing the furniture around it.
Describe the new object’s material and approximate size.
“Keep the amber pump bottle. Place it on pale travertine against a warm cream studio background with soft window light.”
A different setting changes the presentation while keeping the bottle as the subject.
Name the surface, backdrop, and light direction.
“Extend this coastal photo into a wide landscape. Continue the sea, rocky shore, and sky around the lighthouse.”
The wider composition adds a coastline and open water beyond the narrow frame.
Choose extension when you need new space; cropping only keeps part of an existing image.
“Create a sage ceramic perfume bottle with a wooden cap on pale limestone. Add soft olive-leaf shadows and warm afternoon light. No text or logos.”
A written brief sets the subject, material, composition, and lighting without an uploaded image.
Include the subject, setting, and lighting before adding small details.
Keep working from what works
The canvas holds your source and its alternatives together. Move images aside, zoom in to inspect an edge, and select the version you want to continue. You do not have to turn your latest result into the only surviving copy.
Upload the main photo first. Extra images can guide appearance or style. Describe which object to change and what should remain: “Make the chair dark green; keep the wood legs and the room.” Mark an area when several similar objects would make the instruction ambiguous.
Choose Auto or a listed ratio, then select a resolution. The canvas shows the task price before you send it. There is one output per image task. You can leave the page after acceptance; background processing and internal review continue without keeping your browser open.
The new image arrives separately. Inspect the requested change and its boundaries before exporting. To explore another direction, select an earlier version and describe the next edit. Your canvas keeps that branch alongside the newer work instead of replacing it.
Portraits, products, places, and ideas from scratch. Open an image to see the prompt, then make it your own.
Choose a monthly allowance or a one-time credit pack. Save 50% with annual billing; subscription credits still arrive each month.
Start with 30 welcome credits, awarded once per eligible account identity. Subscription credits arrive monthly, including on annual plans, and expire at the end of their credit month. Buying a pack keeps free-account storage at 1 GB and removes AI export watermarks.
What AI Photo Editor saves, what a task costs, and where a little judgment still matters.
Upload a photo and use crop, rotate, flip, resize, or format conversion without an account or credits. These basic edits run in your browser. You can also prepare an AI request before signing in; the image, prompt, and settings stay with your draft through login. AI Photo Editor asks you to sign in before a paid image task starts, so there is an account to save the result and track its credit charge.
Select the image you want to change and describe the object, its location, and the intended result. You can mark a region when the target needs more precision. AI Photo Editor uses the surrounding picture as context and places the edited region back into its source. Pixels outside that editing boundary are retained. Edges still need inspection, especially around fine hair, reflections, or overlapping objects. Adding something beyond the frame can extend the image automatically.
One image task has one quoted price and produces one result. You can ask for several related changes to that image together; the number of objects mentioned does not become a separate bill. The default Auto resolution uses 1K and costs 10 credits. Choosing 2K costs 20 and 4K costs 30. A cutout or solid background task costs 4. The canvas shows the price before you confirm. A new task to try another variation has its own quote.
We check submitted requests and review generated images before delivery. A clear mismatch, such as making a requested blue garment red, can trigger an internal correction without a second credit charge. Technical recovery is also our responsibility. If a task cannot finish, its reserved credits are released according to their original expiry. A delivered image that follows your description may still differ from your taste; choosing to generate another version is a new paid task.
The resolution labels select the image service’s output tier. They do not guarantee a particular width and height for every ratio. In our square-output testing, the 4K tier produced 2880 × 2880 pixels, rather than 4096 × 4096. AI Photo Editor shows the actual dimensions of each image on the canvas. Choose the size for your intended use, then check the exported file before sending it to a printer or a platform with strict pixel requirements.
Saved projects and uploaded files belong to your account; other visitors cannot browse them. An AI task sends the required images and instructions to our processing providers, so cloud editing is different from a local browser-only crop. Free-account AI exports have a watermark. Once your account has made a qualifying subscription or credit-pack purchase, AI exports have no watermark. Basic local edits do not use the AI watermark. See the privacy policy for processing and retention details.
An accepted AI task continues in the background after you leave. Return to the project to check its status and result. Each completed image is kept as a separate version, so a later change does not replace the original or erase the result before it. Select an older version to start a different branch. Signed-in projects save automatically; wait for the saved indicator before relying on another device. Guest drafts are temporary browser data, so download anything you want to keep.
Use a JPG, PNG, WebP, or supported HEIC image up to 20 MB. Export the selected image as JPG, PNG, or WebP; use PNG or WebP when you need transparency. You must have permission to upload and edit the material, and requests are checked against our content rules. Ask the agent an image-editing question through the free consultation option, or use Contact for account and payment questions. Include a task reference when available, but never send passwords or API keys.
Completion time varies with the request, output resolution and service demand. We do not promise a fixed turnaround. Check the task status on the canvas; an accepted task continues if you leave the page. If processing takes longer than expected, reopen the same project before starting a duplicate request. Contact support with the task reference if you need help.
Check that you have the necessary rights to the source images, references, people, logos and other material in the result. A paid plan or credit pack removes the AI export watermark; it does not establish rights to third-party content or guarantee that an output is unique or suitable for your intended use. Review our Terms and inspect product details before publishing an image in a listing or advertisement.
You can select reference images on the canvas and describe what each should contribute. A task accepts up to 16 input images in total, including the selected source, and returns one result. Reference images guide that result; they are not a batch-processing queue. To edit several source photos, submit separate tasks and check each displayed credit price.
Saved project files remain until you delete the project or account, subject to service operation and storage limits. Deleted projects have a seven-day recovery window and still count toward storage during that time. After the window, cleanup removes files no remaining project or version uses. Clearing browser data only removes local drafts; it does not delete a saved project. The Privacy notice explains provider processing and separately retained payment records.
Your prompt stays in place. Return to it, add the image you want to work on, and open the AI Photo Editor canvas when you are ready.