Does the AI background remover send my photo to a server?

No. The model weights are downloaded to your browser once, and inference runs locally using WebGPU where it is available and WebAssembly where it is not. That is why the first run takes longer than the second, and why the tool keeps working offline afterwards.

Which format gives the smallest file?

For photographs, AVIF is usually smallest, WebP close behind, and JPEG largest of the three at equal visual quality, but support and encode time run in the opposite order. WebP is the safe default in 2026: it is supported everywhere that matters and is meaningfully smaller than JPEG.

Is image metadata removed when I convert a file?

Re-encoding through the canvas drops EXIF data, which includes the camera model and, on phone photos, often the GPS coordinates where the picture was taken. That is a side effect worth knowing about in both directions: it protects you when sharing, and it loses data you may have wanted to keep.

Can I process several images at once?

The batch tools accept multiple files and apply the same operation to each, which is the right approach for a folder of photos that all need the same resize or the same format. Processing happens one image at a time in the tab, so a large batch is bounded by your device rather than by a queue.

20 tools · runs in your browser

Image Tools

Compress, convert, resize, crop and edit images on the canvas in your own browser.

Files never leave your device
No sign-up
Free, no watermarks

All Image Tools

Grouped by what you want to get done. Star a tool to find it again on the homepage.

Compress and resize · Smaller files and exact dimensions, one image or a batch.

Image Compressor

Compress JPEG, PNG, WebP & AVIF images with quality control, no upload needed

Resize Image

Resize images to exact pixel dimensions or by percentage - all in browser

Batch Resize

Resize hundreds of images at once to the same dimensions

Crop Image

Crop any image to a precise selection with aspect ratio locking

Edit and enhance · Rotate, flip, sharpen, adjust colour and remove backgrounds.

Image Editor

Full in-browser image editor: crop, resize, rotate, flip, adjust colors, blur, annotate and export

Background Remover

Remove backgrounds from images using WebGPU/WASM AI models, no upload

Brightness / Contrast

Adjust the brightness, contrast, saturation and hue of any image

Sharpen Image

Sharpen blurry photos and enhance image details using browser filters

Rotate Image

Rotate images by any angle to fix a sideways photo or straighten a horizon

Flip Image

Mirror images horizontally or vertically

Convert and encode · Change format, make icons and embed images as Base64.

Convert Image Formats

Convert images to PNG, JPG or WebP, one at a time or in a batch

SVG Converter

Convert SVG to PNG/JPG or rasterize SVG files at any resolution

ICO Generator

Create multi-size ICO favicon files from any image

Image to Base64

Encode any image file to a Base64 data URL string

Base64 to Image

Decode a Base64 data URL string back into an image file

Draw and create · Draw and annotate, add a watermark, or make a QR code.

Watermark Image

Add a text watermark to images, in a corner, the centre or tiled across

Drawing Board

Full-screen drawing app: sketch on a blank canvas or over a photo with brushes, shapes, text and layers

QR Code Generator

Generate QR codes for links, Wi-Fi, contact cards and UPI payments

Read and inspect · Pull text out of a picture and check its hidden metadata.

OCR - Image to Text

Extract text from any image using Tesseract.js OCR running in browser

Remove Metadata

Strip EXIF and other metadata from images for privacy

Often used alongside these, from other categories.

About these Image Tools

Almost every image problem is one of three questions: which format, how many pixels, and how much quality to throw away. Getting those three right is what separates a 180 KB hero image from a 4 MB one that looks identical. These tools run on the browser's own canvas and codecs, which means the picture (a passport scan, a photograph of a person, an internal dashboard) is decoded and re-encoded on your machine and nowhere else.

Choosing between these Image Tools

Resize before you compress, never after

A 4000-pixel-wide photo displayed in a 800-pixel column is carrying five times the pixels anyone will see. Cutting dimensions removes data that compression then does not have to encode at all. Doing it the other way round means compressing detail you are about to discard.

Format follows content, not preference

Photographs belong in JPEG, WebP or AVIF, where lossy compression works with continuous tone. Screenshots, logos and anything with flat colour and sharp edges belong in PNG or WebP, where JPEG artefacts around text are immediately visible. Transparency rules out plain JPEG entirely.

Every re-encode is permanent

Lossy formats do not remember what they discarded. Opening a JPEG, editing it and saving it again applies the loss a second time, and it compounds. Keep an original, work from it, and export once, rather than repeatedly editing the export.

Background removal is a model, not a measurement

The AI background remover runs a segmentation model with WebGPU or WebAssembly on your device. It is good on clear subject-background separation and imperfect on hair, glass and motion blur. Check the edges before you use the result somewhere that matters.

Questions people ask

Does the AI background remover send my photo to a server?: No. The model weights are downloaded to your browser once, and inference runs locally using WebGPU where it is available and WebAssembly where it is not. That is why the first run takes longer than the second, and why the tool keeps working offline afterwards.
Which format gives the smallest file?: For photographs, AVIF is usually smallest, WebP close behind, and JPEG largest of the three at equal visual quality, but support and encode time run in the opposite order. WebP is the safe default in 2026: it is supported everywhere that matters and is meaningfully smaller than JPEG.
Is image metadata removed when I convert a file?: Re-encoding through the canvas drops EXIF data, which includes the camera model and, on phone photos, often the GPS coordinates where the picture was taken. That is a side effect worth knowing about in both directions: it protects you when sharing, and it loses data you may have wanted to keep.
Can I process several images at once?: The batch tools accept multiple files and apply the same operation to each, which is the right approach for a folder of photos that all need the same resize or the same format. Processing happens one image at a time in the tab, so a large batch is bounded by your device rather than by a queue.

Guides on this topic

Camera RAW files, and the JPG hidden inside every oneA RAW file is not a photo. It is the sensor's measurements, waiting to be developed. But every camera also tucks a finished JPG inside, and that is often all you need.iPhone photos in HEIC: opening, sharing and converting themYou send a photo from your iPhone and the other person gets a file they cannot open. The photo is fine. The format is the problem, and it has three fixes.Base64 images: when inlining helps and when it costs youA data URI embeds the picture in the markup. That is occasionally the right call and usually a way to make a page slower.Screenshots that actually help: annotation, redaction and what leaksA black box over text in a screenshot really does delete it, unlike in a PDF. Blurring, on the other hand, has been reversed.