Remove text from images
Give your image a fresh start. Upload a photo, screenshot or illustration, remove unwanted text and review the result before saving.
Make room for the image
Upload a source, choose the type of text to remove, then compare the result before downloading.
Source image
Upload an image with text
Drop a JPG, PNG, or WebP here, up to 16 MB.
Clean result
Your cleaned image appears here.
Choose a model
Cleanup mode
300 credits per generation. Sign in and use your account credits.
Try a source image
Images below are source examples and use-case illustrations. Generate a cleanup with your own image to see the result.
How to remove text from an image
1Upload your image
Start with a JPG, PNG or WebP up to 16 MB, or choose a sample to explore the workspace. Sign in before uploading your own image.
2Choose what to clean
Choose a mode for captions, posters, comic lettering or scan marks. Select a model and add optional notes about details you want to keep, then select Remove text.
3Review and download
Compare the source and result. Zoom in to check lettering, edges and repaired textures, then use the download button to save the generated image.
Give product images another use
An old promotion or caption can date an otherwise useful product shot. Start with Poster cleanup to remove the typography, then check product edges, packaging and shadows before reusing the image. AI may also change nearby details.
Prepare screenshots for sharing
Remove visible names, captions or notes from an image before adding it to a presentation. Inspect every area you intended to clear. AI cleanup is a visual edit, not a secure redaction tool, and it does not remove information from your original file.
Reuse the visual, rethink the message
Keep a useful composition in play when campaign copy changes. Clear old headings or labels from your own graphic, download the image and add the new message in your editor. Text sitting over faces or detailed patterns may need extra cleanup.
Put the artwork at the center
Prepare your drawings and comic panels for a portfolio by clearing draft lettering or annotations. Comic panel cleanup asks the model to keep borders and line art. Compare the result with your source so the final image still reflects the work you want to show.
Frequently asked questions
How do I remove text from an image?
Upload your image, choose a cleanup mode and select Remove text. You can add notes about details to preserve. Review the generated result beside the source and download it when it works for your use.
Which image files can I upload?
Upload one JPG, PNG or WebP image up to 16 MB. You need to sign in to upload your own file and generate a result.
Is the text remover free?
Generation uses account credits. Basic costs 300 credits, Pro Fast costs 600 credits and Pro costs 1200 credits per generation. Pro Fast and Pro require a paid plan. Your available balance and plan determine which options you can use.
Will everything behind the text stay exactly the same?
No. The model generates a repair for the area covered by text, so it cannot recover the hidden original pixels. It may change nearby textures, edges or small details. Inspect the result at full size before using it.
Can I remove just one caption and keep other words?
Use the cleanup mode closest to your task and add a note naming the caption to remove and the text to keep. These are instructions to the model rather than a precise selection mask, so some text you wanted to keep may also change.
Can it remove lettering from comic panels?
Choose Comic panel cleanup for speech bubble text, sound effects and annotations. The request asks the model to keep panel borders and character line art, but closely overlapping lettering and artwork may need additional editing.
Can I use it to hide personal information?
It can help remove visible text from a screenshot or photo, but it is not a secure redaction tool. Check the downloaded image carefully and use a dedicated redaction workflow for sensitive information. Your original file remains unchanged.
How do I save the cleaned image?
After generation, open the result at full size to inspect it, then select its download button. The tool saves the generated image file; it does not promise the same file format or pixel dimensions as the upload.