Social Media Image Resizer

This tool turns a photo into the exact pixel size a social platform asks for, directly in your browser. It is for a YouTube thumbnail, an Instagram post or Story, a Facebook post or Page cover, a LinkedIn post or banner, an X post or header, a Pinterest pin or a TikTok cover — the twelve sizes listed on this page, picked from a list instead of typed in.

Two things have to happen when a picture goes into a fixed frame, and the tool asks you to choose which one. Fit keeps the whole picture: the image is scaled until it fits inside the frame and the leftover space is filled with the background colour you choose, so nothing is lost — a square photo in a 16:9 thumbnail gets bars at the sides. Fill covers the frame: the picture is scaled until it fills the frame completely and the edges that overflow are cropped away, so there are no bars but part of the picture is cut off. A 4:3 photo filled into a square loses a strip from each side.

In Fill mode you can move the crop with two focus sliders (across and down, as a percentage). The crop is centred on that point and then slid back inside the picture if it would run past an edge, so a focus point near a border parks the crop as far that way as it can go. This is a manual fraction, not face detection: put the point roughly over the subject and the crop follows. In Fit mode the focus sliders are hidden because nothing is cropped.

Output is JPG by default at quality 90 — the safest format for platforms, since every one of them accepts JPG. PNG keeps sharp logos and screenshots (and its alpha channel in Fill mode), and WebP makes a smaller file. In Fit mode the frame is filled with the background colour first, so the saved image is opaque and the picture's own transparent areas show that colour too; choose Fill with PNG output if you need transparency kept. If the frame is larger than your picture, the result is enlarged by interpolation — more pixels, not more detail — and the tool says so on that result.

How it works: your browser decodes each image (using its EXIF orientation, so phone photos are upright), draws the chosen part onto a canvas at exactly the preset's width and height, and re-encodes it. When an image shrinks by more than half it is scaled in halving steps with the browser's high-quality smoothing, so edges stay clean. EXIF metadata — camera, date, GPS location — and embedded colour profiles are not copied to the result.

These are the sizes the tool used on its last update. Platforms change their recommended dimensions, and they crop and re-compress uploads again on their own side, so treat the preset as a close starting point and confirm the current requirement in the platform's own help centre (linked below) before something with a deadline. If you only need different pixel dimensions, use the image resizer; if the file is too large, use the image compressor.

Limits: up to 50 images at once, 50 MB each, and any single output is well under a browser's canvas ceiling (16,384 px per side, 100 megapixels). Everything runs on your device — the Network tab in your browser's developer tools shows no request carrying your files.

1. The file is size-checked (50 MB) and its leading bytes confirm it is PNG, JPG, WebP, GIF, BMP or AVIF, whatever its name says. 2. The preset's frame is fixed at the width and height in the table below, e.g. 1280 × 720 for a YouTube thumbnail. 3. Fit — where the whole picture goes: scale = the smaller of (frame width ÷ image width) and (frame height ÷ image height); the drawn size is image size × scale, and it is centred: x = (frame width − drawn width) ÷ 2, y = (frame height − drawn height) ÷ 2. The frame is filled with the background colour before the picture is drawn. 4. Fill — which part of the picture is kept: the crop takes the frame's aspect ratio. If the image is wider than the frame, the crop keeps the full height and its width is height × (frame width ÷ frame height); otherwise it keeps the full width and its height is width × (frame height ÷ frame width). The crop is placed at x = focusX × image width − crop width ÷ 2 (and likewise for y), then clamped to 0 … image size − crop size so it stays inside the picture. 5. The chosen region is drawn onto a canvas of exactly the preset's size, in halving steps when it is shrinking by more than 2×, then encoded as JPG, PNG or WebP at the quality you pick (90 by default). 6. The file is named after the original plus the preset and its size: logo.png → logo-youtube-thumbnail-1280x720.jpg.