Bulk Image Resizer

How do you resize a mixed batch to a useful maximum size?

Choose Max edge, set Longest side (px) to 1200, and click the count-based Resize button. This keeps landscape and portrait proportions while limiting their longest side. For a batch of three photos, the button reads Resize 3 images; the completed files are available individually or through ZIP (3).

One batch with a 1200px longest-side limit

Original dimensions	Scale factor	Output dimensions
4000 × 3000	1200 ÷ 4000 = 0.3	1200 × 900
2000 × 3000	1200 ÷ 3000 = 0.4	800 × 1200
800 × 600	No enlargement	800 × 600

The calculation is scale = limit ÷ longest side, followed by rounding each scaled dimension to the nearest whole pixel, with a minimum of one pixel. Sources already within the limit retain their dimensions. This makes Max edge useful for mixed photo collections where a common bounding size matters more than identical frames.

When should you choose Size, Percent or Max edge?

Percent applies the same percentage to each source's own width and height. The quick buttons are 25%, 50% and 75%, with a custom percentage field alongside them. At 50%, each dimension halves, so the pixel count becomes one quarter of the original. That arithmetic does not predict the compressed file size: format, content and encoder quality also affect bytes.

Size starts at Custom size with target fields of 1200 × 800 and the aspect ratio locked. With the lock active, processing uses the target width and calculates height separately from each image's original ratio. The displayed height follows the first image when width changes; it is not a promise that every differently shaped source will have that height. Unlock the ratio when both target dimensions must apply.

Max edge starts at 1200 and offers quick limits of 640, 800, 1200, 1600 and 2048 pixels. It never enlarges an image. Choose Image Resizer for a single-image workflow rather than managing a batch.

How do Fit, Fill and Stretch change the result?

These options appear in Size when the aspect ratio is unlocked. For a 4000 × 3000 source going into a 1200 × 1200 square, the differences are concrete:

• Fill uses the central 3000 × 3000 source region, removing 500 pixels from each horizontal edge before scaling. It fills the square without changing proportions, but subjects near the sides can be clipped.
• Fit scales the complete photo to 1200 × 900 and centers it on a white 1200 × 1200 canvas. The top and bottom padding are 150 pixels each. It preserves the full frame but adds a visible background.
• Stretch draws the whole source into 1200 × 1200, changing its proportions. It avoids cropping and padding at the cost of distortion.

Fill is centered automatically; there is no subject selection or movable crop anchor here. Use Bulk Image Cropper when cropping is the main task. Fit explicitly paints white behind the image, so choose it deliberately for transparent artwork.

Which social and web presets are built in?

The Preset dropdown contains ten Social media entries and six Web / Screen entries. These are fixed shortcuts in this tool, not a live check of platform publishing requirements. Selecting one sets both dimensions, unlocks the aspect ratio and selects Fill.

Preset label	Width × height in pixels
IG Post / IG Portrait / IG Story	1080 × 1080 / 1080 × 1350 / 1080 × 1920
FB Post / FB Cover	1200 × 630 / 851 × 315
X / Twitter / YouTube Thumb	1200 × 675 / 1280 × 720
LinkedIn / Pinterest / TikTok	1200 × 627 / 1000 × 1500 / 1080 × 1920
HD / Full HD / 2K / 4K	1280 × 720 / 1920 × 1080 / 2560 × 1440 / 3840 × 2160
Thumbnail / OG Image	300 × 200 / 1200 × 630

The orientation swap control exchanges target width and height across the batch; it does not rotate source pixels. Square targets do not need a swap. For rotating the actual pictures, use Bulk Image Rotate.

Why can dimensions stay unchanged or still grow?

Do not upscale smaller images starts checked. In Percent it prevents percentages above 100 from enlarging sources. In Size it retains original dimensions only when both original axes fit within the calculated target. A mixed case with one axis shrinking and the other growing still uses the requested target. The checkbox therefore does not prohibit every possible increase along one axis.

The unchanged badge means equal pixel dimensions, not an untouched file. Every processed item is decoded, drawn and encoded again, including items whose dimensions stayed the same. Output quality, format and metadata can therefore differ. A yellow banner appears when completed results contain no dimension changes.

What do output format and quality actually control?

Output format offers Original Format, JPEG, PNG, WebP and AVIF. Original Format resolves each source's format separately; it does not preserve the original compressed bytes. Quality starts at 85. For JPEG, WebP and AVIF, that value becomes 0.85 for the browser encoder. PNG export omits the quality parameter and uses lossless encoding of the resulting pixels, not lossless preservation of the source.

Encoding uses the browser's Canvas API rather than a bundled format encoder, so requested formats depend on browser encoding support. The tool does not verify the returned blob's type against the requested type. Check downloaded output when using a format your browser may not encode. There is no target-KB setting or guaranteed percentage reduction; use Bulk Image Compressor for a compression-focused workflow.

What should you check before downloading the batch?

The upload hint lists up to 50 JPEG, PNG, WebP or AVIF files. Treat that as the intended batch size rather than a memory guarantee: decoded images, previews and the ZIP use device memory. Reduce the batch if large originals strain your device. Processing runs sequentially, and a failed item displays its error while the loop continues.

Resizing uses createImageBitmap(), a local canvas with high-quality smoothing enabled, and toBlob(). It sends no source images to a server and does not copy source EXIF metadata into the output. There is no metadata-preservation switch or guaranteed interpolation algorithm beyond the browser's smoothing setting.

Click a completed thumbnail for the before/after preview, review dimensions and byte sizes, and edit the filename stem if needed. Names initially end in _resized; the batch downloads as resized-images.zip. After changing settings, use the count-based Re-resize All button before downloading. Reprocessing starts from original files, avoiding a chain of repeated resizes of previous outputs.