Bulk resize product images for any marketplace
Bulk resizing is not one operation, it is a decision about which file is the source of truth. Keep one master image per product at the largest size you will ever need, and derive every marketplace size from it. Resize once in the wrong direction and the damage is permanent.
The mechanical part is easy: pick the target size, apply it to a folder, download a ZIP. The part that saves real time is the order — largest target first, padding instead of stretching, and a naming pattern you can still read a month later.
Opens the tool with no preset. Pick a platform and rule first, then drop your images in.
Why every marketplace needs a different file
Each platform publishes its own main image size, and they do not agree. Amazon wants 1000×1000, Walmart goes to 2000×2000, eBay sits at 1600×1600, and Shopify stores accept up to 2048×2048. Shorter-form marketplaces such as TikTok Shop list 800×800.
Uploading the same file everywhere is the mistake to avoid. Too small and a platform upscales it or rejects it. Too large and it is accepted but heavier than it needs to be, which costs you page weight on the storefront.
| Marketplace | Main image size | Background | Formats |
|---|---|---|---|
| Amazon | 1000×1000 | pure white (RGB 255, 255, 255) | JPG / PNG |
| Walmart | 2000×2000 | pure white (RGB 255, 255, 255) | JPG / PNG |
| eBay | 1600×1600 | any background | JPG / PNG |
| Shopify | 2048×2048 | any background | JPG / PNG |
| TikTok Shop | 800×800 | pure white (RGB 255, 255, 255) | JPG / PNG |
| Shopee | 800×800 | any background | JPG / JPEG / PNG |
| AliExpress | 800×800 | pure white (RGB 255, 255, 255) | JPG / JPEG / PNG |
| Temu | 800×800 | pure white (RGB 255, 255, 255) | JPG / PNG |
Not every spec is square
Square is the default shape for a main image, but it is not universal. Walmart also lists a portrait main image at 1500×2000, and SHEIN lists 800×1000, which is taller than it is wide.
That is why "resize to 1000×1000" is not a workflow, it is a habit that breaks on the first portrait spec. Read the target size and aspect per platform, and let the tool pad instead of stretch. Padding keeps the product proportions honest; stretching makes a bottle look like a different bottle.
If a source image is landscape and the target is portrait, a padded result is correct: the product sits centered with white on the left and right. Never distort the product to fill the frame.
The one-source-many-sizes workflow
- Pick the master file: Choose the largest good version you have per product. If the source is smaller than the biggest target you need, reshoot rather than upscale.
- Fix the background once: Cut out and place on white, or keep the original scene if every target accepts any background. Do this on the master, so all derived sizes inherit the decision.
- Sort targets from largest to smallest: Generate 2048×2048 first, then 2000×2000, then down to the 800 px class. Each step steps down from a larger source, so nothing is ever upscaled.
- Run one folder per target: Batch by target size, not by product. One run for Amazon, one for Walmart, one for the 800 px platforms. That keeps output folders unambiguous.
- Check a sample before shipping the batch: Open two or three outputs at 100%: confirm exact pixel dimensions, white corners, and no visible halo around the product edge.
- Name and store the outputs: Keep the target size or platform in the folder name. Six months from now, "walmart-2000" tells you what a file is; "final-final-2" does not.
Padding, stretching, and cropping: pick two
A resize has to reconcile a source aspect with a target aspect, and there are only three honest ways to do it. Stretching fits the frame and distorts the product. Cropping fills the frame and loses product edges. Padding keeps the product intact and adds background.
For marketplace main images, padding is almost always the right answer, because the product silhouette is the thing being judged. Crop only when the source has generous empty margin around a centered product, and never stretch.
A padded result is also easier to spot-check: white bars on two sides mean the aspect differs from the target, which is expected behavior, not a bug.
Skip the upload, not just the clicking
Most bulk resizers are web services: you upload the folder, wait, then download a ZIP. That is fine for a handful of mockups and less fine for a catalog, where the same files also contain unreleased products and supplier pricing sheets.
The batch image tool inverts that: the images are decoded, resized, and re-encoded inside your browser tab. Nothing is uploaded, so a hundred-photo run costs you no bandwidth in either direction, and the processing keeps working after you disconnect.
The free tier covers 5 images per batch, which is enough to validate the output on a real folder before you commit to a full catalog run. Paid plans lift that batch limit: US$4 per month or US$12 per year.
What bulk resizing cannot fix
- Soft or blurry sources. Resizing down preserves detail; it never creates it.
- A background that is not pure white on platforms that require pure white (RGB 255, 255, 255). Resize does not change pixel values in the background.
- The wrong crop. If the product touches the frame edge in the source, every derived size inherits that crop.
- Missing secondary images. A batch job produces one file per source; it does not invent detail shots or lifestyle scenes.
Start with one batch
Pick a single marketplace, run five images through the tool, and compare the output against the spec line for that platform before you touch the rest of the catalog. Once the folder comes back correct, the only remaining variable is volume.
Frequently asked questions
Can I resize many images at once without uploading them?
Yes. The tool runs entirely in the browser: images are read from your disk, processed in memory, and written back as a download. No server receives them, so there is no upload step and no queue.
Should I upscale a small image to reach 1000×1000?
No. Upscaling cannot recover detail that was never captured, and the result looks soft on the product page. Reshoot or find a larger original instead.
Does resizing change the aspect ratio?
Only if you stretch. Padding keeps the source aspect and fills the remaining area with background, which is what marketplace main images expect. Stretching is available in many editors and is almost always wrong for product photos.
What format should the batch output use?
Use the format the platform lists. For Amazon that is JPG / PNG. When a platform accepts several, JPG is the safe default for photographic product images.
How many images can I process in one batch?
The free tier handles 5 images per batch. Paid plans raise the batch size; pricing is on the pricing page.