EAN / GTIN Product Image Extractor

Catalog, PIM, and marketplace teams paste EAN-8, UPC-A, EAN-13, or GTIN-14 barcodes and receive primary plus gallery product image URLs, with product name and brand when public catalogs expose them. Each row includes source provenance and a collection timestamp so agents can trust what they store. Built for enrichment workflows on Apify with scheduling, API, and MCP access.

The Actor is HTTP-first against free Open Facts catalogs (food, beauty, pet, and general packaged goods) with a bounded UPCitemdb Explorer free-plan fallback. It validates GS1 check digits, deduplicates codes, and only persists real HTTPS image URLs.

Best fit for this Actor

• Choose this Actor when you already hold retail barcodes and need image URLs plus light identity fields for PIM, listing QA, or agent enrichment.
• It works well for food, beauty, pet, and many packaged-goods codes covered by Open Facts, with UPCitemdb helping on broader retail UPCs.
• For Amazon listing research after you have an ASIN, start with Amazon Product Intelligence, then pass confirmed barcodes here for image galleries.
• For supplier pack-shot context from customs filings, pair it with ImportYeti Scraper when your workflow needs importer/supplier evidence next to barcode images.

PIM backfill workflow

A catalog manager exports a list of retail EAN-13 codes from their PIM. They paste Nutella (3017620422003) and LIFEWTR (012000161155) into Barcode codes, keep max images at 10, and run. The dataset returns primaryImageUrl, imageUrls, productName, brand, and source for each found code. They export JSON into the PIM image field and schedule the same input weekly so new SKUs pick up fresh public gallery URLs.

Quick start input

{
  "codes": ["3017620422003", "012000161155"],
  "maxImagesPerProduct": 10,
  "concurrency": 2
}

Open the Input tab in Console for every field. Prefill already includes the two codes above so a first run stays cheap.

Input reference

Field	What it controls
codes	EAN-8 / UPC-A / EAN-13 / GTIN-14 strings (spaces and dashes stripped). Max 100 unique valid codes.
maxImagesPerProduct	Gallery URL cap per barcode (1–20). Primary image is always imageUrls[0] when present.
maxCodes	Hard cap after dedupe for quality tests.
concurrency	Parallel lookups (1–8). Keep 1–3 for free-catalog courtesy.
sources	Optional filter: openfoodfacts, openbeautyfacts, openpetfoodfacts, openproductsfacts, upcitemdb.
includeEmptyRows	When true, write unbilled NOT_FOUND rows for valid codes with zero images.

What data you receive

One dataset item is one barcode that resolved to at least one real image URL (unless you enable empty rows).

{
  "ean": "3017620422003",
  "gtin": "3017620422003",
  "normalizedGtin": "03017620422003",
  "format": "EAN-13",
  "productName": "Nutella",
  "brand": "Nutella, Ferrero",
  "imageUrls": [
    "https://images.openfoodfacts.org/images/products/301/762/042/2003/front_en.879.400.jpg"
  ],
  "primaryImageUrl": "https://images.openfoodfacts.org/images/products/301/762/042/2003/front_en.879.400.jpg",
  "imageCount": 1,
  "source": "openfoodfacts",
  "confidence": 0.88,
  "scrapedAt": "ISO-8601 collection time"
}

Field	Description
ean / gtin	Validated barcode digits
normalizedGtin	Zero-padded GTIN-14 for dedupe
productName / brand	From the winning source when present
primaryImageUrl / imageUrls	Real HTTPS image URLs only
source / confidence	Provenance plus heuristic confidence
scrapedAt	Collection timestamp for the row

Download the dataset as JSON, CSV, Excel, or HTML from the run Output tab.

Pricing

Startup billing and empty runs

The apify-actor-start event costs $0.00005 per startup unit. Apify charges one startup unit per GB of allocated memory, with a minimum of one unit, so a 2 GB run incurs two units. This startup charge still applies when input is invalid, no results are found, or an upstream request fails. Statements about free, unbilled, or skipped results refer only to the corresponding result event; they do not waive the startup charge. Apify platform usage is included in the event prices and is not passed through as an additional customer charge. Check the live Pricing tab for current events, tiers, and rates. Result limits cap result events, not the entire run cost.

This Actor uses Pay per event with platform usage included. Check the live Pricing tab for the current schedule.

Event	Typical price	When charged
apify-actor-start	$0.00005	Once per run
image-set-found	$0.008	Once per barcode that returns at least one image

Worked example: two found barcodes in a run allocated at most 1 GB ≈ $0.00005 + 2 × $0.008 event charge, with Apify platform usage included. Invalid checksums and codes with no images skip the image-set-found event.

Outcomes and agent readback

Outcome	Meaning
COMPLETE	Every valid code returned images
PARTIAL	Mix of found and not-found or upstream gaps
VALID_EMPTY	Valid codes, no images found
INVALID_INPUT	Empty input or every code missed checksum/format validation
UPSTREAM_FAILED	Providers returned transport issues with no successful miss responses

Read OUTPUT / RUN_SUMMARY for outcome, itemsPushed, and chargedImageSets. Provenance lives on each row in source and scrapedAt.

API / MCP example

curl "https://api.apify.com/v2/acts/khadinakbar~ean-gtin-image-extractor/runs?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"codes":["3017620422003"],"maxImagesPerProduct":5,"concurrency":1}'

On Apify MCP this Actor appears as apify--ean-gtin-image-extractor. Pass codes, then read dataset rows for primaryImageUrl and provenance.

Agent prompt card

Resolve these product barcodes to image galleries. Use ean-gtin-image-extractor with codes. Expect one dataset row per found barcode with primaryImageUrl, imageUrls, source, and collection time. Treat missing codes as VALID_EMPTY scope, and confirm cost against the Pricing tab before large batches.

Best results guidance

• Provide canonical retail barcodes (EAN-13 / UPC-A) rather than internal SKUs.
• Start with Open Facts sources for food, beauty, and pet packs; confirm electronics and general retail with upcitemdb when needed.
• Keep concurrency low and validate checksums in your own sheet first so billed rows stay productive.
• Schedule focused batches when catalogs refresh so agents always see current public image URLs.

Builder's note

I designed this Actor for image-set resolution with provenance rather than nutrition dumps or paid per-file downloads. In my testing, a checksum gate plus a free multi-source waterfall kept MCP payloads flat and honest: found codes charge once, unknown codes finish empty, and every saved URL is a real HTTPS link from the winning catalog.

Legal / responsible use

Use only for legitimate catalog, research, and enrichment workflows under applicable laws and site terms for the public catalogs you query. Product images remain copyright of their owners; this Actor returns public URLs and grants no redistribution rights.

Open Food Facts and related Open Facts projects are trademarks of their owners. This independent Actor is not affiliated with, associated with, or endorsed by Open Food Facts. UPCitemdb is a trademark of its owner. This independent Actor is not affiliated with, associated with, or endorsed by UPCitemdb. GS1 is a trademark of GS1. This Actor is not affiliated with or endorsed by GS1.

Support

Open an Issue on the Actor page with the barcode and RUN_SUMMARY.outcome when a result needs review.