The hard specs: what Temu checks on upload
These are the requirements that gate publication. Everything else in this guide is optimization; this section is pass/fail.
| Requirement | Spec | Enforcement |
|---|---|---|
| Minimum images | 3 to publish | Blocks listing |
| Maximum gallery | Up to 10 total | Soft cap |
| Main image background | Pure white (255,255,255) | Algorithmic check |
| Minimum resolution | 800 × 800 px | Blocks upload |
| File formats | JPG, JPEG, PNG | Blocks upload |
| File size | Under 3 MB per image | Blocks upload |
| Aspect ratio | Square (1:1) for main | Strongly preferred |
| Size chart (apparel) | Uploaded as an image, not a table | Category-gated |
Two of these deserve extra attention. The first is resolution: 800 × 800 is the floor, not the target. Temu's mobile app serves zoom on the main image, and an 800 px file falls apart the moment a shopper pinches in. Shoot and export at 1600 × 1600 or higher and let the platform downsample.
The second is the apparel size chart. Unlike Amazon or Shopify, where sizing lives in structured fields, Temu expects an uploaded image in the size chart slot. That means it has to be legible at thumbnail scale on a phone, and it should carry both centimeters and inches — a large share of Temu's catalog crosses regions, and a chart in inches only will generate returns from buyers who guessed.
Marketplace specs move. Confirm current numbers in Temu Seller Center for your specific category before processing a thousand SKUs against them — category-level rules (jewelry, apparel, electronics) frequently override the general defaults.
Why "white background" is stricter here than you think
Most rejected main images are not obviously wrong. They are photographs taken on white paper, in a white light tent, that a human reviewer would call white without hesitation. The problem is that cameras do not record backdrops as pure white — they record them as whatever the lighting, white balance, and paper color produced.
Sample any untouched studio shot and you will typically find the background sitting somewhere between #F2F2F2 and #FAF8F5. That is a few percent off pure white. To a shopper it is invisible. To a validator comparing pixel values against 255,255,255, it is a fail — and because every SKU on that shoot shares the same setup, it fails as a batch.
The fix is a background replacement, not a brightness slider. Pushing exposure until the backdrop clips to white also blows out the product's highlights — white shirts lose their seams, chrome loses its edge definition, and pale packaging turns into a silhouette. Cutting the product out and compositing it onto a true 255,255,255 canvas keeps the product's own tonal range intact while making the background unambiguous.
Retouchable handles that as a batch operation: run the catalog through background removal, output onto pure white, and the entire shoot clears the validator together instead of one SKU at a time.
Before uploading, open a main image and use the eyedropper on four corners plus the area directly beside the product. If any sample reads anything other than 255,255,255, the file is not ready — regardless of how white it looks on your monitor.
The 10-slot gallery order that sells on a price-led marketplace
Temu shoppers browse fast, compare hard, and buy on perceived value. The gallery is where a listing wins or loses that comparison, and slot order matters more than image count. A tight six-image gallery in the right sequence outperforms ten images in a random order.
| Slot | Image | Job it does |
|---|---|---|
| 1 | Product on pure white | Passes review, earns the click |
| 2 | Lifestyle / in-use | Establishes context and desirability |
| 3 | Scale reference | Prevents the #1 return reason |
| 4–5 | Material close-ups | Answers "is this cheap?" |
| 6 | What's in the box | Sets delivery expectations |
| 7 | Size chart (apparel) | Cuts fit-related returns |
| 8 | Variant / colorway grid | Reduces bounce to competitors |
| 9 | USP or comparison graphic | Differentiates in a crowded category |
| 10 | Alternate angle or detail | Fills remaining doubt |
Slot 3 is the one most sellers skip and most regret. Temu's catalog skews toward products where buyers cannot intuit dimensions from a floating white-background photo, and a listing that omits scale generates a predictable stream of "much smaller than expected" returns. An in-hand shot, a shot beside a common object, or a dimensioned overlay on a white-background image all work.
Slot 9 earns its place only in categories where you are visibly competing on the same search results page as ten near-identical products. If your differentiation is real — thicker material, more pieces, a longer cable — a simple side-by-side graphic converts. If it is invented, it reads as noise.
The rejection triggers worth memorizing
Beyond background color, a short list of content rules accounts for most of the remaining first-pass failures. None are surprising, but all are easy to inherit from images built for another channel.
Gets rejected
- Watermarks, logos, or seller branding on the main image
- Promotional text: "50% OFF", "Free Shipping", "Best Seller"
- Borders, frames, or decorative edges
- Collages or multiple products in the main slot
- Placeholder or stock supplier imagery reused verbatim
- Accessories shown but not included in the purchase
Passes clean
- Single product, centered, filling most of the frame
- Pure white background with no visible seam or gradient
- Natural shadow or no shadow — both are acceptable
- Text confined to secondary slots where it aids comprehension
- Original imagery you own the rights to use
- Only what ships in the box, shown accurately
The reused-supplier-imagery problem deserves a specific warning. If you dropship, the manufacturer's photo pack is almost certainly already live on the listings of a dozen competitors. That is not a review failure in itself, but it guarantees you look identical to everyone underselling you. Re-shooting or regenerating the imagery is one of the few levers available to a seller who cannot win on price.
Adapting an existing catalog without reshooting
Most sellers arriving on Temu already have images — built for Shopify, Amazon, or a wholesale line sheet. The work is conversion, not creation, and it splits into four mechanical passes you can run across the whole catalog.
1. Normalize the background. Cut every main image out of its original backdrop and place it on a true white canvas. This single pass resolves the largest rejection category and also makes the catalog visually consistent, which matters when a shopper browses several of your SKUs in one session.
2. Square and center. Product photos built for Shopify are often 4:5 or 3:2. Re-canvas to 1:1 with the product centered and roughly 85% of the frame filled. Do not stretch — pad with white.
3. Strip overlays. Line sheet images frequently carry SKU codes, brand marks, or price stamps burned in. These have to come off the main image; anything genuinely useful can move to a secondary slot.
4. Fill the gaps. Audit which of the ten slots you can populate from existing assets, then identify what is missing. In practice it is almost always the scale reference and the size chart — the two images nobody shoots because they are not "hero" images and the two that most reliably reduce returns.
Run the background normalization pass first and re-upload a small test batch of five to ten SKUs before processing the full catalog. If a category-specific rule is going to bite, you want to discover it on ten listings rather than a thousand.
Handled manually, a 500-SKU conversion is weeks of retouching work at prices that make the exercise hard to justify on thin marketplace margins — professional retouching commonly runs $25–50 per image, which puts a single catalog pass into five figures. Handled as an automated batch, it is a queue you start and check on later, which is the difference between adding a sales channel and postponing it indefinitely.