Describe the change and an AI model makes it: a new background, an object removed, a cartoon or painting style. Sign in and describe your edit. ToolGPT uploads the photo and shows the settings and cost before you start.
1
Sign in with Google or Apple, then drop a JPEG, PNG or WEBP above.
2
Say what to change and what to keep. ToolGPT uploads the selected photo and shows a plan with the quality and credit cost.
3
Confirm, and the edited picture arrives with a download button, usually within a minute.
Some changes take an expert hours with a brush: replacing a busy background, taking a stranger out of a holiday photo, turning a portrait into a watercolor, bringing back the colour of an old print. The AI photo editor does them from a sentence, such as “replace the sky with a sunset, keep the people as they are”.
This tool works differently from the free photo tools. The photo is sent to our server and on to an outside image service running an OpenAI model when you start the requested edit. Each edit costs credits: the estimate is shown first, the credits are held while the model works and refunded if it fails. The result is square, 3:2 or 2:3, whichever is closest to your photo, and as a new drawing it can differ in small details.
Yes. AI editing needs you to sign in with Google or Apple; the free photo tools do not.
It is paid in credits, and the estimate is shown before you confirm. Higher quality costs more. Credits are refunded if the edit fails.
Yes. Requesting an edit uploads the selected photo to prepare the plan. Starting the edit processes it on a server.
Close to it: the result is square, 3:2 landscape or 2:3 portrait, whichever is nearest.