Image to Text (OCR)
Turn screenshots, scanned pages and photos of printed text into editable text without sending the image anywhere.
Cut the subject out of a photo and save it as a transparent PNG, processed entirely on your own device.
Removing a background by hand with selection tools can take a long time, especially around hair or soft edges. This tool does it automatically: choose a photo, and a segmentation model separates the main subject from everything behind it, leaving you with a PNG file whose background is transparent.
Your photo is handled inside the browser and never uploaded. That matters for portraits, family pictures and unreleased product shots. The model behind it is released under the permissive Apache 2.0 license.
The tool runs MODNet (Xenova/modnet) through Transformers.js. MODNet is a matting network originally designed for portraits: for every pixel in a resized copy of your image it estimates how much that pixel belongs to the foreground, from fully opaque to fully transparent. That estimate, called an alpha matte, is scaled back to the original size and applied as the transparency channel of the output image. Because the matte has soft values rather than a hard cut, edges such as hair fade out more naturally than a simple outline would. MODNet is published under the Apache 2.0 license.
MODNet is released under the Apache 2.0 license, which allows commercial use. You are still responsible for having the rights to the photo itself, especially pictures of other people.
PNG supports transparency. JPEG does not, so a JPEG would fill the removed area with a solid colour.
No. The image and the mask are processed and combined locally in your browser.
Turn screenshots, scanned pages and photos of printed text into editable text without sending the image anywhere.
Find common objects in a photo and draw labeled boxes around them, using a detection model that runs locally.
Get a short, plain-English sentence describing what appears in a photo, generated entirely on your own device.
Transcribe English speech from an audio file or your microphone using a Whisper model that runs on your own device.
Condense a long English article or report into a few sentences, generated by a model that runs on your device.
Paste English text and see whether each line reads as positive or negative, with a confidence score for each.