Image Collage Maker

Image Collage Maker composes multi-photo layouts inside a single 2D Canvas. Each layout slot is computed as a rectangle from the chosen output size and gap, then images are drawn with cover-fit logic — the same algorithm CSS object-fit: cover uses, where the source is centred and scaled so that one axis fills the cell while the other overflows and is clipped.

Three layout families are available. Grid uses a fixed column count (2, 3, or 4) and lays cells out left-to-right, top-to-bottom. Strip places every image in a single horizontal or vertical strip. Feature gives the first image a 60 percent hero cell and stacks the remaining images in a sidebar — useful for product hero shots with thumbnails. Auto picks columns as the integer square root of your image count, which produces visually balanced grids for 4, 9, or 16 images.

Output presets cover the dimensions that platforms actually want: Square (1200x1200) for Instagram feed posts, Wide (1600x900) for blog covers and Twitter cards, Landscape (1920x1080) for slide and YouTube end-card embeds, and Story (1080x1920) for Instagram and TikTok vertical content. The Canvas is drawn at full output resolution so the export is sharp at the requested size with no upscaling.

Background colour is configurable, gap (the gutter between cells) is a slider, and image order can be rearranged with left/right arrows in the source tray. Each rearrangement re-renders the Canvas immediately, so you can preview multiple orderings in seconds without re-uploading.

Export goes through Canvas toBlob in your choice of WebP, JPEG, or PNG. WebP and JPEG honour the quality slider; PNG is lossless and is the right pick if your collage includes screenshots or graphics with hard edges that JPEG would smear into blocky artefacts.

Up to 24 images per collage is the practical cap. Going higher would either make every cell too small to read on a phone or force more memory than typical mobile browsers can hold while keeping every source bitmap decoded simultaneously.

All decoding, drawing, and encoding happens locally. The source files live as object URLs; the Canvas exists only in the tab; the final Blob is downloaded directly. There is no upload step and no server-side compositing pipeline — closing the tab releases everything.