“Add,” “edit” and “replace” can sound interchangeable, but they lead to different image operations. Choosing the right one prevents a new caption from being sent through an unnecessary reconstruction workflow—or old lettering from remaining underneath an overlay.
What are the four image-text operations?
| Intent | What happens |
|---|---|
| Add | Draw new movable text above the picture. |
| Edit | Broadly update words already embedded in the pixels. |
| Replace | Specify an exact old phrase and its new phrase. |
| Remove | Erase selected words and reconstruct the area without new wording. |
Who needs this decision guide?
Creators, product teams, marketers and support writers use it when they have only a JPG, PNG or WebP and need to decide whether the visible text should remain, change or disappear.
Choose Add for a new layer
Use Add Text to Image for a caption, label, watermark or headline that never existed in the source. You can drag and style the layer and export PNG.
Choose Edit or Replace for embedded words
Use the main editor for a broad existing-text change. Use Replace Text in Image when the instruction is explicitly “old phrase becomes new phrase.” Both rely on AI reconstruction and require review; neither recovers the original editable text object.
Choose Remove when nothing should replace it
The Remove Text workflow rebuilds the selected area without inserting new words. Only modify images you own or are authorized to change.
What should you do next?
State the desired final image in one sentence. If it includes “new words above,” choose Add; “these old words become,” choose Replace; “update existing wording,” choose Edit; and “leave no wording,” choose Remove.
FAQ
Can an overlay hide existing words?
It may cover them visually, but it does not reconstruct the background and is not the same as replacement.
Which workflow is free of AI credits?
The current Add Text overlay workflow does not use credits; sign-in is required for its PNG download.