YourTools.
简体中文
Back to the catalog
Tools

Upscale images with AI

Enlarge photos 2× or 4× with Real-ESRGAN, an AI model that rebuilds sharp detail instead of just stretching pixels. It runs in your browser, in seconds with WebGPU, for photos up to 4 megapixels.

Process locally

Files stay in this browser session. Closing or clearing releases them.

Choose files

JPG, PNG, WebP, BMP, TIFF, up to 5

Drop files here, or

Adjust settings

Optional: Enlarge

Upscale images with AI

Run

Processes in this browser; cancel at any time

Choose files first

About this tool

Enlarge JPG, PNG, WebP, BMP and TIFF photos 2× or 4× with Real-ESRGAN, an AI model that sharpens edges and removes JPEG blockiness and some noise instead of just stretching the pixels: for example to print a small photo larger, improve an old picture from a phone or a website, or get a bigger image for a shop listing. The photos are processed in your browser and never uploaded.

The model, 4.9 MB, downloads the first time together with the ONNX Runtime engine that runs it, and the browser then keeps both. With WebGPU, which current Chrome and Edge offer, a photo takes seconds; otherwise the processor does the work, which is much slower: about a minute per megapixel of the original on a fast desktop and several minutes on a laptop. Up to 5 photos of up to 25 MB and 4 megapixels each, 80 MB in total, go in one run.

How to use it

  1. Choose or drop up to 5 JPG, PNG, WebP, BMP or TIFF photos of up to 4 megapixels each.
  2. Pick how much to “Enlarge”: “2× (twice as wide and tall)”, the default, or “4×”.
  3. Select Run. The first run downloads the model; later runs start right away.
  4. Check the results in the preview, then download them one by one or all together as a ZIP. Each keeps its format and file name, transparent areas stay transparent, and JPG and WebP are saved at quality 0.92.

Formats and limits

Input
JPG, PNG, WebP, BMP, TIFF, up to 5 files
Output
Same format as the input
Size limit
25 MB per file, 80 MB in total
Image size limit
4 megapixels per image
Settings
Enlarge (2×, 4×)
Privacy
Files stay in the current browser session. No file bytes are uploaded.
Price
Free, no sign-up

Common questions

Are my photos uploaded?

No. Only the model is downloaded from this site; the photos are enlarged in your browser and stay on your device.

How long does it take?

With WebGPU, a photo takes a few seconds. Without it, the processor does the work: about a minute per megapixel of the original on a fast desktop, several minutes on a laptop. The network always enlarges 4 times and 2× is that result scaled down by half, so 2× takes as long as 4×. A run may take at most 30 minutes, not counting the first download, so on a slow computer enlarge one or two photos at a time.

How large can the photos be?

The original may have up to 4 megapixels, such as 2000 × 2000 or 2560 × 1440, and the result up to 36 megapixels. So 4× takes photos of up to 2.25 megapixels, and a 1920 × 1080 picture becomes 7680 × 4320; photos between 2.25 and 4 megapixels can be enlarged 2×. Make larger photos smaller first, for example with “Resize images”.

Can it fix a blurry photo?

Only in part. It makes edges sharper and removes JPEG blockiness and some noise, but it cannot recover detail that was never captured: it may invent plausible texture instead, and faces and small text can look odd. It works best on photos; illustrations work too.

Which model is it, and may I use the results commercially?

Real-ESRGAN’s general model, realesr-general-x4v3, blended half and half with its weak-denoise variant, as Real-ESRGAN’s own script does by default. It is released under the BSD-3-Clause licence, which allows commercial use. Its licence and source are listed in this site’s third-party notices.

Local first

Adjust image brightness, contrast and saturation

Change the brightness, contrast, saturation, hue, warmth and sharpness of photos, or make them black and white (grayscale), sepia or negative. The same settings apply to every image in the batch.

Use tool
Local first

Blur faces in photos

Find every face in your photos and cover it with pixels (mosaic), a blur or a black shape, to hide who is in a picture before you share it. Faces are found by a small model that runs in your browser.

Use tool
Local first

Compress images

Reduce image file sizes with adjustable quality and a target file size.

Use tool
Local first

Convert image format (HEIC to JPG, PNG, WebP)

Convert images between JPEG, PNG, WebP, GIF, BMP, TIFF and ICO icons, and turn iPhone HEIC or HEIF photos into JPG, PNG or WebP; AVIF can also be read where the browser decodes it. Animated images are not supported, and files are not uploaded.

Use tool
Local first

Crop images

Crop one region from every image, in pixels or percent, as a rectangle, a circle or ellipse, or a rounded rectangle. Outside a shape is transparent, so JPEG and BMP shapes are saved as PNG.

Use tool
Local first

Split images into a grid

Cut every image into rows × columns tiles, such as a 3 × 3 grid for social posts, optionally from the centred square. Tiles keep the input format and are numbered by row and column.

Use tool