Combine Multiple Images into One

Side by side puts the images in one row, left to right. It suits two or three pictures of a similar shape: a before and after, two versions of a design, or two photos in one frame for a post. Beyond three or four, a row gets very wide and each picture ends up small on a phone screen.

Stacked puts them in a column, top to bottom. That is the natural layout for screenshots of a conversation, the steps of a tutorial or a long page captured in parts. Portrait photos stacked on top of each other get very tall, so for those a row or a grid usually reads better.

A grid arranges four or more images in rows and columns. Left on Auto, it keeps up to three images in one row, puts four in a 2×2 grid, nine in 3×3 and twenty in five columns of four rows. Choose 2 to 6 columns yourself for a particular shape. When the last row is not full, Last row decides whether its images sit on the left, in the center or on the right.

The quick-start presets are shortcuts to these layouts. 2 side by side and 3 in a row set up a row at matching heights, 2×2 grid sets two columns of equal cells, and Before & after adds Before and After labels to a pair. A preset changes the arrangement only, so your background, spacing and label wording stay as they were.

Photos rarely arrive at the same size, and lining up a 4000-pixel photo with a 1200-pixel screenshot by hand is tedious. In a row, Same height scales every image to one height so the top and bottom edges line up; when stacked, Same width does the same for the sides. Proportions are always kept, so nothing is stretched.

Scale to decides which height or width they share. Smallest image is the default: everything is reduced to match the smallest picture, so nothing is enlarged and every image stays sharp. Largest image enlarges the smaller pictures instead, which gives a bigger result but can make them look soft.

Original size leaves every image at its own pixel size. The space beside a shorter image is filled with the background, and Align places the image at the top, middle or bottom of the row, or on the left, center or right of a stack. It is the right choice for screenshots, where any scaling blurs the text.

In a grid every cell has the same shape. Auto takes the shape of the median image, so a single panorama among portraits does not decide it, while 1:1, 4:5 and 16:9 force square, portrait or widescreen cells. Fill and crop covers each cell and trims whatever overflows; Fit whole image keeps every picture complete and leaves background around those that don’t match the cell. Cell size from works like Scale to, taking the cell size from the smallest or the largest image.

Switch Mode to Overlay to layer images instead of placing them next to each other. The result is the size of the base image, and any part of a layer that hangs over its edges is cut off.

1. Add the background picture first, then the images that go on top. The first image in the list is the base and the next five become Layer 1 to Layer 5; later layers sit in front of earlier ones, so reorder the thumbnails to change which is on top.

2. Drag a layer on the preview to place it. It snaps to the edges and the center of the base and shows a guide line when it does; hold Alt while dragging to place it freely. On a touchscreen, drag it with your finger.

3. Set its Size as a share of the base’s width, from 1% to 200%, turn it with Rotation and fade it with Opacity. To move it by a single pixel, select it and press the arrow keys; with Shift, each press moves it 10 pixels.

4. Choose a blend mode if the layer should mix with the picture beneath it rather than cover it, then download. Reset layer puts the selected layer back where it started.

Opacity sets how much of a layer shows. At 100% it covers the base completely; at 50% the two pictures mix evenly. Give a layer the same shape as the base, set its size to 100% and lower the opacity, and one image fades into the other in a double exposure. The fade is the same across the whole layer: gradient fades are not part of this tool.

Blend modes mix a layer with what lies under it, using the standard formulas behind the modes of the same name in most photo editors. Multiply and Screen are the two worth knowing. Multiply drops white, so a signature or a logo scanned on white paper sits on a photo with no white box around it. Screen drops black, so fireworks, flares or stars shot against a black sky blend in by themselves. The table below compares all six.

Logo or watermark places the selected layer in a corner or the center, at a fifth of the base’s width and 70% opacity, with a small margin from the edges: the usual setup for a watermark. A logo saved as a transparent PNG works best, because only the mark shows. A JPG logo has a solid background that converting it to PNG does not remove, so remove the background first, or set the layer to Multiply if the logo is dark on white.

This tool layers images, not text. To type a text watermark on a picture, use the Photo Editor, and to stamp every page of a document, add a watermark to the PDF.

Background fills the gaps, the border and any space beside images that don’t fill their place. White and black take one click, Custom color takes any color, and Transparent leaves those areas empty. In the preview, a checkerboard marks the transparent pixels.

Transparency inside your images is kept as well, so a PNG logo or a cut-out stays see-through wherever it goes. Rounded corners on a transparent background give the result truly rounded edges, which looks tidy on a slide or a web page.

PNG is lossless and keeps transparency, which suits screenshots, graphics and anything with a transparent background. JPG gives the smallest file for photos but cannot store transparency, so transparent areas turn white. WebP keeps transparency and is usually smaller than PNG. The tool starts with PNG when the result has transparent areas or any of the images is not a photo, and with JPG when they are all photos; you can switch either way.

An animated GIF is added as its first frame, because the result is a single still image. iPhone HEIC photos and AVIF files open like any other picture, and the combined image downloads as PNG, JPG or WebP.

The preview uses reduced copies so it can redraw instantly while you adjust things. The download is drawn again from your original files at full resolution, so the preview’s lower resolution never reaches the file.

With Smallest image, the default, pictures are only ever scaled down, which keeps them sharp, and Original size does not scale them at all. What costs sharpness is enlarging, with Largest image or a layer set larger than its own size, because a browser cannot add detail that was never in the file.

JPG and WebP are saved at quality 92 unless you change it, which is hard to tell apart from the original in a photo; the slider runs from 50 to 100. PNG is always lossless. If the file is too heavy for where it is going, run it through the image compressor afterwards, or resize it to the exact dimensions a site asks for.

Very large results are capped so the browser does not run out of memory: 40 megapixels or 16,384 pixels on a side on a computer, and about 16.7 megapixels on phones and tablets. A bigger layout is scaled down as a whole to fit, and the tool says so and shows the size it would have had. Photos are turned upright using their EXIF orientation, and because the file is new, camera details and GPS location are not copied into it.

Bring the images to the same shape before you combine them. Two 4:3 photos line up perfectly side by side, while a 4:3 photo next to a 9:16 one leaves empty space or gets cropped. Crop Image can lock both to the same aspect ratio first.

For a before and after, keep the framing identical, put the original first and use Same height with labels on. A gap of 10 to 20 pixels in white or black reads as a deliberate divider, while no gap makes the pair look like one picture.

Gaps, borders and rounded corners are measured in pixels of the final image. A 20-pixel gap looks thin between two 4000-pixel photos and thick between two small screenshots, so judge it in the preview rather than by the number.

To put a person or a product on a new background, cut it out with the background remover, then add the new background first and the cut-out second in Overlay. For a plain color backdrop, make a blank image at the size you need and add it first. Before you post the result, check its pixel dimensions and file size with the image size checker.