Meta catalog image requirements: the exact specs
Meta applies these rules to the image_link field in a product feed, whether you upload via a data feed, a partner platform integration like Shopify, the Commerce Manager UI, or the Catalog API.
| Requirement | Minimum | Recommended |
|---|---|---|
| Dimensions (Shops / commerce) | 500 x 500 px | 1024 x 1024 px |
| Dimensions (catalog ads) | 500 x 500 px | 1200 x 1200 px or larger |
| Aspect ratio | Accepts non-square | 1:1 square |
| File format | JPG or PNG | JPG for photos, PNG for transparency |
| Main image | 1 required (image_link) | 1 clean, on-white or on-neutral shot |
| Additional images | Optional | Up to 10 via additional_image_link |
| Image URL | Must resolve publicly | Stable CDN URL, HTTPS, no auth |
Two of these deserve emphasis because they are where real catalogs break.
Square is not required, but everything is displayed square. Meta will ingest a 1200 x 1600 portrait image. It will then center-crop it to 1:1 for the Shops grid, the collection ad tile, and most Advantage+ placements. If your product sits in the top third of a tall frame — common with apparel shot on a model, or bottles shot vertically — the crop cuts the product in half. Submitting square source images removes the guesswork entirely.
The URL matters as much as the pixels. Every image_link must return a valid image to Meta's crawler on demand, not just to a browser with your session cookie. The most common causes of "image could not be fetched" errors are hotlink protection on the CDN, a robots.txt rule blocking Meta's user agent, URLs behind a staging password, and query-string image transforms that expire.
Meta formally retired the hard 20% text-overlay limit for ads years ago, and plenty of outdated guides still cite it as a rejection trigger. It no longer blocks delivery. But heavy text overlay still reduces performance, and for catalog images specifically, promotional overlay ("50% OFF", "FREE SHIPPING") violates Meta's commerce product-image policy and can get items rejected from Shops outright. Keep price and promotion claims in the copy fields, not baked into the pixels.
What Meta actually rejects: the commerce policy side
Technical specs get you ingested. Commerce policy gets you approved for Shops. These are separate reviews, and the second one is where sellers lose items without understanding why.
Meta's product image policy for commerce surfaces requires that the image show the item being sold, accurately, and on its own. The recurring rejection causes:
- Promotional text or badges in the image. Sale stickers, "best seller" flashes, discount callouts, shipping claims. This is the single most common commerce-policy image rejection.
- Collage or multi-product main images. A grid of four colorways as the main image reads as a bundle that doesn't match the single-item listing. Use
additional_image_linkfor variants, or split them into separate items with a shareditem_group_id. - Placeholder or "coming soon" graphics. Meta treats a listing whose image isn't the product as a mismatch between image and description.
- Cropped-to-illegible products. If a reviewer can't tell what's being sold, the item fails. Leave breathing room — roughly 5–10% margin on all four sides.
- Watermarks, logos, and borders. Brand watermarks across the product are flagged. A small logo on the product itself is fine; a semi-transparent brand mark over the whole frame is not.
- Regulated-category framing. Supplements, alcohol, and health products face additional restrictions on how they can be depicted, including implied health claims made visually.
Before a large catalog upload, run 10 representative items through as a test batch and let them fully process. Commerce policy review usually resolves within 24–48 hours. Fixing a systemic composition problem across 10 items is trivial; fixing it across 4,000 after a bulk rejection is a project.
Main image vs. additional images: how to use all ten slots
Most catalogs populate image_link and stop. That leaves the ten additional_image_link slots empty — and those slots are what fill the product detail view when a shopper taps through from a Shops tile or a catalog ad.
A sequence that works across categories:
| Slot | Shot | Job it does |
|---|---|---|
image_link | Clean product, square, neutral background | Wins the tap in a crowded grid |
| Additional 1 | Lifestyle / in-context | Establishes use case and aspiration |
| Additional 2 | Scale reference (in hand, in room) | Prevents size-driven returns |
| Additional 3 | Detail / texture macro | Substitutes for touching the product |
| Additional 4 | Alternate angle or back | Answers "what does the other side look like" |
| Additional 5 | Colorway or variant range | Surfaces options without a click |
| Additional 6–10 | Packaging, what's included, care/spec card | Closes remaining objections |
The ordering logic is that the main image is competing for attention, and every additional image is answering a question. That distinction is why lifestyle shots generally belong in slot one of the additional images rather than as the main image: in a grid of thumbnails, a busy scene loses to a clean product, but on the detail page a clean product on white tells a shopper nothing new.
Vendors and feed platforms report meaningful engagement lifts on Instagram Shopping for listings that pair a clean main image with lifestyle support imagery rather than relying on plain product shots alone. Treat specific percentage claims from feed tools with appropriate skepticism — they're not from Meta — but the directional finding is consistent enough across sources, and cheap enough to test on your own catalog, that it's worth structuring for.
Bringing an existing catalog into compliance
The realistic starting point for most sellers isn't a fresh shoot — it's a few thousand existing images of inconsistent size, background, and crop, already live on a Shopify or WooCommerce store and syncing into Commerce Manager. Auditing and fixing that is a sequenced job.
Step 1: Pull the diagnostics Meta already gives you. Commerce Manager's Catalog → Data Sources view lists errors and warnings per item. Export it. Image errors cluster into three buckets: fetch failures (URL problem), dimension failures (too small), and format failures (unsupported type, often a WebP or AVIF served where Meta expects JPG/PNG).
Step 2: Fix the URL and format layer first. This is pure plumbing and usually resolves the largest error count for the least effort. Check that your image CDN serves JPG/PNG to Meta's crawler rather than negotiating WebP by Accept header, that hotlink protection allowlists Meta, and that no image URLs carry expiring signatures.
Step 3: Triage on resolution. Anything under 500 px is dead — it needs a new source file or AI upscaling. Anything between 500 and 1024 px is functional but underpowered; upscaling recovers usable detail for retail-grade display without a reshoot. Anything at or above 1024 px square is fine.
Step 4: Standardize the crop and background. This is the step that visibly changes conversion, and the one most catalogs skip. Inconsistent backgrounds and product scale across a Shops grid read as an unmaintained store.
Manual cleanup
- Per-image retouching, typically $25–50 per image at agency rates
- Days-to-weeks turnaround on a mid-size catalog
- Consistency depends on which retoucher got which batch
- Recrops for a new placement mean paying again
AI batch processing
- A fraction of traditional retouching cost per image
- Whole catalog processed in a single pass
- One background and crop standard applied uniformly
- Re-export to new aspect ratios on demand
Batch background removal, square-canvas normalization, and consistent margin padding are exactly the kind of high-volume, low-judgment work AI tooling handles well — and it's why platforms like Retouchable exist: standardizing an entire catalog to a single spec is a mechanical transformation, not a creative one.
Step 5: Backfill additional images. Once the main images are compliant and consistent, populate the additional slots. This is where the remaining upside sits, and it requires no changes to items already approved.
Meta vs. other marketplaces: what carries over
If you already meet Amazon or Google Shopping image standards, you are most of the way to Meta compliance — but not all the way, and the gaps are specific.
| Rule | Meta | Amazon | Google Shopping |
|---|---|---|---|
| Minimum dimension | 500 px | 500 px (1600 px for zoom) | 100 px (250 px apparel) |
| Recommended | 1024–1200 px sq | 2000 px longest side | 800 px+ |
| Pure white background required | No | Yes (main image) | No |
| Promotional text allowed | No | No | No |
| Watermarks allowed | No | No | No |
| WebP accepted | No — JPG/PNG | No | Yes |
| Additional image cap | 10 | 8 (varies by category) | 10 |
Two practical consequences. First, an Amazon-ready image set is Meta-safe on background — Meta doesn't require pure white, so a white-background Amazon main image passes fine, and often looks cleaner in a Shops grid than a busy alternative. Second, a Google Shopping feed is not automatically Meta-ready, because Google accepts WebP and much smaller minimums. Sellers who built their image pipeline around Google's looser rules routinely hit format and dimension failures the first time they push to Commerce Manager.
The efficient approach for multichannel sellers is to master at the strictest common denominator — a 2000 px square JPG with a clean neutral background and no overlay — and derive every channel's requirements from that single source file. Downscaling to Meta's 1024 px is free. Recovering detail you never captured is not.