Why your imported AliExpress images look so bad
If you dropship or resell, the photo problem is rarely the supplier's fault in the way you think. You open a listing you imported through DSers, Zendrop, CJ, or AliDropship, and the main image lands at 400-600px, soft, and visibly compressed. It fails your store's zoom and gets flagged on marketplaces. The instinct is to blame the factory for shooting bad photos — but most of the time the factory's photo was fine, and the damage was introduced downstream by the import pipeline.
Here is the chain that quietly degrades a supplier product image. AliExpress serves every photo through a CDN that generates resized variants on the fly, with size suffixes baked into the URL: _640x640.jpg, _350x350.webp, _220x220.jpg. Import apps and store themes frequently grab one of these smaller variants — not the seller's original upload — and then your store re-encodes it again as JPEG. So a aliexpress image low quality complaint is usually two separate losses stacked: a downscaled CDN variant, recompressed a second time on import. The community has documented this for years on the AliDropship forums, where sellers trace blurry imports back to the resize suffix rather than the source file.
This matters because it changes your first move. Before you reach for any tool, you want to recover the largest file that actually exists. This guide walks the full rescue path: find the real master, ask the supplier for a hi-res pack when the master is still too small, and only then upscale and clean up what you genuinely have. For the broader playbook across every platform, see our hub on how to make product photos clearer.
Step one: recover the supplier's real master file
Do not upscale anything yet. The image your import app handed you is often the worst copy in existence, and a better one is one click away. The technique is to strip the resize suffix from the CDN URL so the server returns the seller's original upload.
- Open the image in a new tab. On the AliExpress product page, right-click the main photo and choose "Open image in new tab." Look at the filename in the address bar — you will usually see a suffix like
...image_640x640.jpgor...image_350x350q90.jpg. - Delete the dimension suffix. Change
image_640x640.jpgto plainimage.jpg(or.webp) and reload. The CDN typically returns the seller's master, which is commonly 1000x1000 to 2000x2000px, and occasionally 3000px or larger. - Right-click and save the full version. That recovered file — not the imported thumbnail — is what you work from.
Repeat this for every gallery image, not just the hero. Gallery shots are often where the largest masters hide because suppliers upload them once and never bother resizing. If you sell at catalog scale, doing this by hand across hundreds of SKUs is brutal, which is where the batch workflow later in this guide earns its keep.
Step two: ask the supplier for a hi-res image pack
Recovering the master only gets you what AliExpress allowed the seller to upload — and that is a real ceiling. AliExpress requires main images of at least 800x800px, recommends 1000x1000px, and generally caps seller uploads around 2000x2000px (per AliExpress seller-spec breakdowns from OreateAI and Deep-Image). In other words, even the supplier's best public file is frequently 800-2000px and was never meant to clear a premium marketplace's bar.
The supplier almost always has more. The studio that shot the product holds source files far larger than anything the AliExpress uploader accepted. A short message works: ask for "the original high-resolution product photos, including any 3000px or larger studio files and any lifestyle shots." Many factories keep a shared drive or a WeChat/WhatsApp album of originals and will send them on request, especially if you are placing repeat orders. This single step can skip the upscaling problem entirely — if the master they send already meets your platform spec, you are done.
When the supplier cannot or will not provide larger files — common with white-label and POD middlemen — then you are working with a genuinely small original, and upscaling becomes the realistic fix rather than a shortcut.
The pixel bar your supplier image has to clear
Before you fix anything, know the target. A 600px supplier thumbnail does not just "look small" — it falls below the hard floor on most marketplaces. Here is where each platform sits as of 2026; cite the official seller-help pages if you need to quote a hard minimum, since these specs shift.
| Platform | Display floor | Zoom unlocks at | Recommended target |
|---|---|---|---|
| Amazon | 500px longest side | 1000px (1600px+ optimal) | 1600px+ longest side (~2000px ideal) |
| Etsy | — | crisp zoom degrades below ~2000px shortest side | 2000px shortest side |
| eBay | 500px longest side | 1600px+ | 1600px+ |
| Shopify | (your own store) | over 800x800px | 2048x2048px |
| Walmart | ~500x500px (auto-unpublishes below) | 1500x1500px | 2200x2200px |
The honest takeaway: a 600px supplier image is below even eBay's and Amazon's 500px display floor, would be auto-unpublished by Walmart Marketplace, and falls far short of Etsy's 2000px-shortest-side recommendation. Amazon frames 1600px+ as the optimal zoom experience (per Jungle Scout's restatement of Seller Central), with 1000px the lowest that still triggers zoom. Etsy recommends at least 2000px on the shortest side — the highest pixel bar of the majors — which is exactly why a recovered 1000px AliExpress master still won't natively pass it. That gap is the whole reason upscaling exists in this workflow. If you sell mainly on Amazon, our deep-dive on increasing image resolution for an Amazon listing covers that platform's spec in detail.
Step three: upscale, denoise, and deblur what you have
Once you have the largest master the supplier can give you and it is still below spec, upscaling is the move — but the right kind of upscaling. A plain resize in Photoshop just stretches the same pixels and makes the softness larger. What you actually need is a tool that does three things at once: enlarges the pixel count, removes JPEG compression noise, and sharpens edges that the recompression smeared.
Snappyit's free Product Photo Upscaler is built for exactly this. Drop in a recovered 1000px master and it upscales toward 4K while denoising the import artifacts and deblurring soft edges — pushing a recovered AliExpress master past Etsy's 2000px line or Walmart's 2200px recommendation in one pass. There is no login, no watermark, and no usage cap, which matters when you are processing a supplier's whole catalog rather than a single hero shot.
The tool is deliberately clarity-only. It will not relight a dim factory shot, recolor a product, or swap a background — those are separate jobs. White-background requirements (which Walmart enforces on main images) belong to a different workflow entirely; if you need that, handle it with a dedicated background tool, not the upscaler. Here, the job is purely resolution and sharpness on a photo you already have.
What upscaling can and cannot do to a crushed thumbnail
This is the part most tool marketing skips, and it is the part that protects you from a wasted afternoon. AI upscalers cannot recover detail that was never captured. If the supplier's original JPEG already destroyed the fabric weave, the fine stitching, or the engraving on a clasp, no model brings it back — it can only make a plausible guess, and past roughly 4x linear enlargement (16x in area) those guesses start inventing structure that looks real but isn't there. The 2026 upscaling guides from Lovart and Snapcorn are blunt about this: a blurry input keeps its blur faithfully, and the best inputs are "small but relatively sharp."
So set the expectation honestly. A crushed 600px thumbnail upscales to something that meets listing pixel specs and looks clean and professional in the gallery — it does not become a studio macro shot. What you reliably get:
- A small but reasonably sharp master → a crisp, spec-compliant listing image. This is the sweet spot.
- A noisy, recompressed copy → a cleaner, larger image with the compression artifacts removed, good enough to pass the gate and stop looking amateur.
- A genuinely blurry or motion-smeared original → bigger, but still soft. Upscaling cannot un-blur what the camera never focused on. This is when you go back and demand the supplier's real source file.
If you are unsure why a specific image is soft in the first place — focus, compression, or downscaling — our guide to fixing blurry product photos has a diagnostic decision tree for picking the right fix.
Step four: batch-process a whole supplier dump to spec
One product is a five-minute job. A supplier sending you a folder of 200 images, or a dropshipping catalog of 500 SKUs, is a different animal — and doing it one file at a time is where the workflow falls apart. This is the real reason resellers give up and ship the bad photos anyway.
The free Product Photo Upscaler handles batches: upload a whole supplier image dump, let it upscale and denoise every file toward your target spec, and download the results as a ZIP. No per-image cap and no login means you can run an entire catalog in a sitting. Process the recovered masters from step one, not the original imports, so each file starts from the best available pixels.
For a step-by-step on organizing and running a full catalog — naming, target sizing per platform, and ZIP handling — see the dedicated guide on batch-upscaling a product photo catalog. If you want to compare batch tools head to head, our roundup of the best free AI image upscalers for e-commerce lays out the options.
Why this is worth doing before you launch a listing
It is tempting to launch with whatever the import app gave you and fix it later. Don't. Dropshipping communities — the AliDropship and AutoDS blogs among them — consistently document that low-resolution and blurry imported images are a direct conversion killer, because online shoppers judge a product almost entirely on its photos before they read a word. A soft, pixelated hero shot reads as "cheap reseller" and sends buyers straight to a competitor selling the identical item with a clean photo.
The compounding compression makes it worse over time, too. Each platform that touches the image re-encodes it, and that recompression "reduces the file size but reduces the quality with it," stacking new artifacts on an already-soft source. Recovering the master and upscaling once, properly, at the start breaks that cycle — you publish the cleanest version that exists rather than the most degraded one. For the strategic view of how photo quality fits the bigger picture of selling, see our pillar on AI product photography.
Frequently Asked Questions
Why are my AliExpress images low quality after importing them to my store?
Almost always because the import app or your store theme pulled a downscaled CDN variant (a URL with a suffix like _640x640.jpg) rather than the supplier's original master, then re-encoded it as JPEG on import. That stacks two quality losses. Recover the master by stripping the size suffix from the image URL before you do anything else.
How do I find the largest version of an AliExpress supplier photo?
Open the image in a new browser tab, look at the filename in the address bar, and delete the dimension suffix — change image_640x640.jpg to image.jpg and reload. The CDN typically returns the seller's original upload, usually 1000x1000 to 2000x2000px and occasionally larger. Save that file, not the imported thumbnail.
Can an upscaler turn a 600px supplier thumbnail into a sharp studio photo?
No. AI upscaling cannot invent detail the original never captured. A small but reasonably sharp image upscales beautifully to listing spec; a crushed, blurry thumbnail becomes larger and cleaner but stays soft. It will meet pixel requirements and look professional in the gallery, but it will not become a studio macro shot.
The supplier's master is only 1000px — is that enough for my marketplace?
It clears eBay and Amazon's display floor and Amazon's 1000px zoom trigger, but it falls short of Etsy's 2000px-shortest-side recommendation, Walmart's 2200px target, and Amazon's optimal 1600px zoom threshold. Upscaling a 1000px master toward 4K is the realistic way to clear those higher bars when the supplier has nothing bigger.
Should I ask the supplier for photos or just upscale what I have?
Ask first. The factory's studio almost always holds source files far larger than what AliExpress allowed them to upload (uploads are capped around 2000x2000px). A short request for the original high-resolution pack can hand you spec-ready images and skip upscaling entirely. Only upscale when the supplier genuinely can't provide larger files.
Can I fix a whole supplier catalog at once instead of one image at a time?
Yes. Snappyit's free Product Photo Upscaler processes batches — upload a full supplier folder, it upscales and denoises every file toward your target spec, and you download a ZIP. There's no per-image cap and no login, so you can run an entire dropshipping catalog in one session.
Does the upscaler also remove the background or whiten it for Walmart's main-image rule?
No, and that's by design. The Product Photo Upscaler is clarity-only — it enlarges, denoises, and sharpens but does not relight, recolor, or change backgrounds. White-background requirements are a separate task handled by a dedicated background tool. Run them as two distinct steps in your workflow.
Will upscaling fix a genuinely blurry or out-of-focus supplier photo?
Only partially. Denoising and deblurring can clean up compression smear and mild softness, but a photo the camera never focused on stays soft when enlarged — there's simply no sharp detail to recover. When the original is truly blurry rather than just small, the right move is to request a better source file from the supplier.
