How to Batch Resize Images: Free Tools and Scripts Compared

Last updated: June 27, 2026

Resizing one image takes a few seconds. Resizing a thousand by hand takes most of a day. The fix is a batch tool: pick a folder, set a target size, run once, and write every output to a new folder so the originals stay safe. This guide covers four reliable, free ways to do it — a browser tool, desktop apps, the command line, and a Python script — with copy-paste commands.

Quick answer: which batch resize method should you use?

Match the tool to how often you do this and how many images you have:

Your situation	Use	Why
One-off, a handful of images, no install	Imagic AI Image Resizer	Browser-based, preset social sizes, no signup
Recurring, hundreds at a time, comfortable in a terminal	ImageMagick mogrify	Fast, scriptable, no upload
Recurring, want a GUI, any OS	XnConvert	Visual actions, unlimited files, free
Automated pipeline or server	Python + Pillow	Full control, integrates into code

If you only do this occasionally, the browser tool is enough. If you resize the same kind of images every week, learn one command-line or script approach and reuse it.

What preset sizes do you actually need?

Social and storefront platforms publish recommended dimensions, and they change over time. Verify the current numbers on each platform's media page, but these are the widely-used defaults:

Target	Dimensions (px)	Aspect	Typical use
Instagram square post	1080 × 1080	1:1	Feed posts
Instagram story / Reel cover	1080 × 1920	9:16	Stories, vertical video
Facebook / LinkedIn link preview	1200 × 630	~1.91:1	Shared link cards
Twitter / X post image	1200 × 675	16:9	In-feed images
YouTube thumbnail	1280 × 720	16:9	Video covers
Ecommerce product (square)	1000–2000 sq	1:1	Marketplace main image
Website hero	1920 wide	varies	Full-width banners

For a full platform reference, see the social media image sizes guide. Always resize to the display size before compressing — it saves more bytes than any quality setting.

Method 1: Browser batch tool (no install)

For a quick job with up to a handful of images, use the Image Resizer:

1. Drag in your images (or select multiple files).
2. Pick a preset size or enter custom dimensions.
3. Choose to preserve aspect ratio (recommended) or force exact dimensions.
4. Run, then download the results.

Best for one-off tasks where installing software or writing a script is overkill. For dozens or hundreds of images, switch to a desktop or command-line method — browser tools meter work and tie up a tab.

Method 2: ImageMagick on the command line

ImageMagick is free, cross-platform, and the fastest way to resize thousands of images. Install it once, then reuse the same command forever.

## macOS
brew install imagemagick
## Windows (Chocolatey)
choco install imagemagick

The key command is mogrify, which operates in place — so always point it at an output folder:

## Resize every JPG to 1920px wide (height auto, aspect ratio kept)
mogrify -path resized -resize 1920x *.jpg

## Resize to 50% of original
mogrify -path resized -resize 50% *.jpg

## Force exact 1920x1080 (distorts — avoid unless you need a fixed box)
mogrify -path resized -resize 1920x1080! *.jpg

mogrify runs single-threaded per image but you can parallelize with xargs:

mkdir -p resized
ls *.jpg | xargs -P 4 -I {} convert {} -resize 1920x resized/{}

On a modern laptop a thousand typical photos finish in well under a minute this way. See the ImageMagick resize docs for the full option list.

Method 3: XnConvert (visual desktop app)

XnConvert is a free GUI batch processor for Windows, macOS, and Linux. It suits people who want to chain actions (resize, then compress, then rename) without writing code.

Steps:

1. Add files or a folder (no count limit).
2. Add a Resize action — by pixels, percentage, or longest side.
3. Add more actions if needed (compress, convert format, rename).
4. Set an output folder and format.
5. Run.

It records your action stack, so a repeat job is one click. Good middle ground between a browser tool and a script.

Method 4: Python script (full control)

For a build step or a server, Python with Pillow gives you total control and no upload. This script resizes every image in a folder to a target width, keeping aspect ratio:

from pathlib import Path
from PIL import Image

input_dir = Path("images")
output_dir = Path("resized")
output_dir.mkdir(exist_ok=True)
target_width = 1920

for img_path in input_dir.glob("*"):
    if img_path.suffix.lower() not in (".jpg", ".jpeg", ".png", ".webp"):
        continue
    with Image.open(img_path) as img:
        ratio = target_width / img.width
        # Skip upscaling — only resize down
        if ratio >= 1:
            img.save(output_dir / img_path.name)
            continue
        new_height = int(img.height * ratio)
        resized = img.resize((target_width, new_height), Image.LANCZOS)
        resized.save(output_dir / img_path.name)

Notes that matter in production:

• Use Image.LANCZOS (or Image.Resampling.LANCZOS in newer Pillow) for downscaling quality.
• Skip images already smaller than the target to avoid upscaling blur.
• Process in parallel with concurrent.futures for large folders.

How do you resize for each common use case?

Use case	Target	Extra step
Ecommerce product photos	1000–2000px square	Add white background + center on canvas
Social media posts	Platform preset (table above)	Export one per platform from one master
Website gallery	Thumbnail 300–400px + full 1600–1920px	Generate both sizes from each original
Email images	≤800px wide, <1 MB	Combine resize with quality compression

Ecommerce square example with ImageMagick — resize then pad to a white square:

mogrify -path out -resize 1000x1000 -background white -gravity center -extent 1000x1000 *.jpg

Best practices

• Always write to a separate output folder. Never mogrify in place without -path — it overwrites originals irreversibly.
• Preserve aspect ratio. Use -resize WIDTHx (height auto). Forcing WIDTHxHEIGHT! distorts the image.
• Test on one image first. Run a single file, inspect it, then batch the rest.
• Only resize down. Upscaling adds no detail and looks soft. Skip files already smaller than the target.
• Resize before compressing. A 4000px image compressed well is still a 4000px download. Combine with the Image Compressor after resizing.
• Keep a master. Store the full-resolution original separately; you can always re-derive sizes from it.

Resizing to the display size is the single biggest byte saving available — bigger than any compression setting. On one real photograph:

A 3648px original shown at 800px on the page downloads ~19× more data than it needs to. No quality setting recovers that — only resizing to the display size does.

Common problems

Symptom	Likely cause	Fix
Output looks blurry	You upscaled	Skip files at or below the target width
Image looks stretched	Forced exact dimensions	Drop the ! — use -resize WIDTHx
Files are still large	Resized but not compressed	Add compression/quality step after resize
Some files error out	Unsupported format or corrupt input	Log and skip; check the format is supported
Originals got overwritten	Ran mogrify in place without -path	Always use -path out; restore from backup

Related guides

• Batch Image Processing Guide
• Social Media Image Sizes Guide
• Compress Images Without Losing Quality
• TinyPNG Alternatives Compared
• Complete Image Optimization Checklist

Frequently asked questions

How many images can a browser batch resizer handle?

Roughly 50 to 200 files per run before a browser tab gets heavy on memory. For larger catalogs, a command-line tool like ImageMagick mogrify handles thousands without the memory ceiling, as the batch processing guide notes.

Does batch resizing reduce quality?

Only if you upscale or pick a poor resampling filter. Downscaling with Lanczos preserves detail; the loss comes from re-compressing, not resizing. Always write to a new folder and keep the originals so any size can be re-exported.

Should I resize before or after compressing?

Resize first. Reducing pixels is the single biggest byte saving, and compressing a full-resolution image wastes effort on pixels nobody will see. The batch processing guide covers the full pipeline order.

Image credits

• Cover — Colorful code on a screen, photo by Nemuel Sereti on Pexels (converted to WebP).
• Resize comparison and byte-savings chart — generated by the author from a landscape photograph (Pexels #1287145, photo by eberhard grossgasteiger) to show real measured file sizes after resizing.