Remove text from an image

Brush over a date stamp, caption or number and the background fills in, on your own device.

Wipe a date stamp, caption, name or number off a picture: brush over the text and an AI fills the background back in, in your browser. It is free, needs no account and adds no watermark, and your files are not uploaded: everything runs in your browser on your device. The first run downloads the object removal model (about 28 MB) and the WebAssembly engine that runs it to your browser, and after that the tool works offline.

Drop the image with text to removeor click to choose a file.
    Details

    Cover the whole object and a little around it, including its shadow. Erase works best on one thing at a time.

    Working…

    Done

    Your files are processed on this device and never sent to a server. Reload the page to clear everything from memory.

    Old digital cameras printed an orange date in the corner of every shot. Memes and screenshots come with captions you want to replace. A photo of a room for a listing shows a house number, a name on a letter or a number plate. Brush over the text, press Erase, and it is replaced with the paper, wall, sky or photo that surrounds it.

    The brush opens small, since text is thin. Paint over each line of letters with a stroke or two that covers their full height. On a plain or evenly textured background the text disappears without a trace; on a busy photo it leaves a soft patch you can refine with another pass.

    If the goal is to hide something before sharing, such as an account number or an address, the fill is a real replacement: the saved file holds only the new pixels. For documents where a visible black box is expected, the blur tool’s Black box setting is the clearer choice.

    How it works

    1. Drop the image.
    2. Brush over each line of text and press Erase.
    3. Press Save; add new text with the Add text to image tool if you need it.

    In your browser, not on a server

    Your browser decodes the photo and you paint the area to remove on a canvas. MI-GAN (MIT, Picsart AI Research), an inpainting network built to run on phones, runs in a web worker through onnxruntime-web (MIT) compiled to WebAssembly. The model crops around the painted area, fills it at 512 × 512 from the surrounding pixels and blends it back at full size; only painted pixels change. The result is saved by your browser’s own image encoder.

    Compared with upload-based tools: Online magic erasers upload your photo to a GPU server, often cap the free size, add a watermark or ask for an account, and keep the upload. The photos people clean up are often personal: family, homes, holidays. Here the network runs on your device and the photo never leaves it.

    Verify it yourself. Give the page a short while after your first run to save itself for offline use, then turn off Wi-Fi or switch to airplane mode and run the tool again. Or open your browser's developer tools (F12, or Cmd-Option-I on a Mac), pick the Network tab, and watch it while you convert. The panel above does the same count for you.

    Questions

    Can someone recover the text I erased?

    No. The painted pixels are replaced with new ones and the file is written fresh, with no layers or edit history, so the original letters are not in it.

    Will it work on a scanned document?

    On a plain page it works well: the text is replaced with blank paper. For a PDF, use Redact PDF, which removes the text itself from the file rather than painting over a picture of it.

    Can it remove a watermark from a stock photo?

    Please do not: a stock watermark marks a picture you have not licensed, and removing it breaches the licence and usually copyright. Buy the licence and download the clean file instead.

    Is the file uploaded anywhere?

    No. There is no server in this process. The page is a static file and the work is done by code running inside your browser tab. Give the page a short while after your first run to save itself for offline use; then you can turn off Wi-Fi and it still works.

    More image tools

    Other conversions

    Guides