Target only the text to erase
Describe the intended wording and preserve the parts of the image that already work.
AI TEXT REMOVAL
Use the removal workspace when the desired result contains no replacement copy. Point to the unwanted caption, annotation, date, or label and ask the editor to rebuild the covered area so it blends with the surrounding photo or graphic.
Edit image textUpload a JPG, PNG, or WebP. Describe the exact text change.

WHAT YOU CAN UPDATE
Use the finished image as context, make the requested copy change, then review the visual result before publishing.
Describe the intended wording and preserve the parts of the image that already work.
Describe the intended wording and preserve the parts of the image that already work.
Describe the intended wording and preserve the parts of the image that already work.
ABOUT THIS IMAGE EDIT
Quote the wording and describe its location, especially when the image contains other text that must remain. A prompt such as “remove the white caption at the bottom, keep the sign in the background” gives the editor a clear boundary.
Mention whether the hidden area contains sky, fabric, a wall texture, a gradient, packaging, or another object. Clean reconstruction is easier when the source image is sharp and the surrounding pattern provides enough visual context.
Zoom in after removal and inspect shadows, straight lines, patterns, and object edges. If the first fill looks uneven, refine it with a local instruction rather than removing additional content. Avoid using the tool to strip ownership marks or watermarks from material you do not control.
Open the image text editorHOW IT WORKS
The editor uses your image and instruction as context. You stay in the conversation until the result is ready.
Start with a JPG, PNG, or WebP poster, photo, menu, screenshot, product visual, or AI-generated graphic.
Say exactly what to replace, remove, or add—and explain what should stay visually unchanged.
Check the result, ask for follow-up changes in the same chat, then download the version you approve.
Review before publishing. Inspect the reconstructed background, edges, patterns, and any nearby text.
COMMON QUESTIONS
The editor attempts to reconstruct the background rather than place a flat block over the words. Results depend on image resolution and how complex the covered area is.
Yes. Name the exact target and say which nearby wording must remain. Location cues are especially helpful when the same phrase appears in multiple places.
Only edit images you own or are authorized to modify. Do not remove copyright, attribution, or ownership marks from third-party content without permission.
MAKE THE EDIT
Open the editor here, upload your visual, and describe the exact result you want.